u~ 8; - Erasmus University Rotterdam · Oncogenes on my mind: ERK and MTOR signaling in cognitive diseases Trends in Genetics, 2008;24(10):498-510 ... Nederlandse samenvatting Dankwoord

~el'.fisehe i3ibliothee'k ErasmusMC

J.. u~ 8; Is- f

COGNITIVE DEFICITS IN CHILDREN WITH NEUROFIBROMATOSIS TYPE 1:

FROM RECOGNITION TO TREATMENT.

COGNITIEVE PROBLEMEN BIJ KINDEREN MET NEUROFIBROMATOSE TYPE 1:

VAN HERKENNING TOT BEHANDELING.

LIANNE CAROLINE KRA.B

The publication of this thesis was financially supported by the N eurofibromatose Vereniging Nederland

ISBN: 978-90-9023507-3

Cover design:

Front- Uanne Caroline Krab, 'The Marvellous Brain', acrylic paint on canvas, SOxSO em (2008).

Back - Blackboard filled with greetings of the children with and without NFl who participated in the

Neurofibromatosis type 1 simvastatin trial and affillated studies.

Printed by PrintPartners Ipskarnp B.V., Enschede

© Lianne Caroline Krab, 2008

All rights reserved. No part of this thesis may be reproduced in any form by print, photoprint, microfllm or any

other means without written permission of the rightful claimant(s). This restriction concerns the entire publication

or any part of it.

COGNITIVE DEFICITS IN CHILDREN WITH NEUROFIBROMATOSIS TYPE 1:

FROM RECOGNITION TO TREATMENT.

COGNITIEVE PROBLE:tvrnN BIJ KINDEREN :tvrnT NEUROFIBRO:MATOSE TYPE 1:

VAN HERKENNING TOT BEHANDELING.

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de

Erasmus Universiteit Rotterdam

op gezag van de

rector magnificus

Prof.dr. S.W.J. Lamberts

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

woensdag 26 november 2008 om 9.45 uur

door

LIANNE CAROLINE KRAB

geboren te V elsen

•' ERASMUS UNIVERSITEIT ROTTERDAM

CONTENTS

CHAPTER!

CHAPTER2

CHAPTER3

CHAPTER4

CHAPTER5

CHAPTER6

CHAPTER7

CHAPTERS

CHAPTER9

CHAPTER10

Introduction, Aims and Outline

Oncogenes on my mind: ERK and MTOR signaling

in cognitive diseases Trends in Genetics, 2008;24(10):498-510

Impact of Neurofibromatosis type 1 on school performance Journal qfChild Neurolo!!J, 2008;23(9): In Press (doi: 10.1177/0883073808316366)

Health Related Quality of Life in children with Neurofibromatosis Type 1: Contribution of demographic factors, disease related factors, and behavior The Journal qf Pediatrics, 2008; In Press (doi: 10.1016 /jjpeds.2008.08.045)

Motor learning in children with Neurofibromatosis type I Submitted (2008)

Quantitative differentiation between healthy and disordered

brain matter in Neurofibromatosis type I patients using Diffusion Tensor Imaging A]NRAmerican Journal qfNeuroradiolo!!J 2008;29(4): 816-22

The NF1 Simvastatin Trial

7.1 The effect of simvastatin on cognitive functioning in children with Neurofibromatosis type 1: a randomized, double-blind, placebo-controlled trial JAMA, 2008;300(3): 287-294

7.2 Challenges for translational trials -Author reply JAMA, 2008; In Press

Discussion and Future Prospects

Summary

Nederlandse samenvatting

Dankwoord

About the author

List of publications

9

33

63

83

101

117

135

137

155

159

177

182

187

193

194

1 0 INTRODL!Cf!ON, _\IMS c\ND OU'IlJNE

Introduction

Over the past few years, mouse models have significantly contributed to our understanding of

the molecular mechanisms underlying cognitive dysfunction in genetic disorders. Moreover,

several preclinical studies in mouse models of for instance Neurofibromatosis type 1 (NF1),

Tuberous Sclerosis Complex, Down syndrome, Rett syndrome, and Fragile X syndrome have

provided evidence that some of these cognitive deficits may be reversible by targeting the

underlying molecular disturbances.l-5 These new findings have sparked a great interest in the

search for drugs that may be used in patients to ameliorate their cognitive problems.6 A recent

study described the beneficial effects of a statin, one of the most widely prescribed classes of

medications, on cognitive deficits of a mouse model for NF1.7 This finding offered an exciting

and unique opportunity to assess the effect of a drug that has been validated in preclinical

studies and for which substantial clinical safety data is available, on cognitive problems in NF1

patients.

This thesis focuses on the recognition and treatment of cognitive problems in children with

NF1. It aims to provide an overview of the specific aspects of cognitive performance that affect

daily life functioning in NF1 children, and tries to identify possible outcome measures that can

be used to assess potential therapeutic interventions. This knowledge was used to perform the

first randomized, double blind, placebo-controlled trial to assess the effect of statins on

cognitive problems in children with NF1.

Neurofibromatosis type 1

Neurofibromatosis type 1 is an autosomal dominant disease with a birth incidence of about 1 in

3000, half of which are sporadic cases.8 It is caused by a heterozygous mutation in the gene

encoding the neurofibromin protein on chromosome 17q11.2.9. IO NF1 patients display

characteristic neurocutaneous abnormalities, such as cafe-au-lait macules and neurofibromas,

and have an increased incidence of malignant tumor formation. NF1 can affect physical

functioning and appearance as well as cognitive performance and behavior.

Clinical manifestations ofNFl

Clinical features of NFl arise predominantly from neural crest derived tissues. The NF1

diagnosis is a clinical diagnosis, based on the presence of two or more major disease features,

such as ccifi-au-lait macules, axillary or inguinal freckling, and neurofibromas (see table 1 and

CH\PllcR 1 11

figure 1).11 Minor disease features of NFl include macrocephaly, hypertelorism, thorax

deformities and small stature.12 The manifestations of NFl develop with age, but usually the

diagnosis can already be made before the age of six.B

Table 1: NIH-defined diagnostic criteria for NF1, with their frequency and typical age of onset

Criterion 11

6 or more cafe-atl-lait macules (>0.5 em in children or >1.5 em in

adults)

2 or more neurofibromas, or

one ple:ciform neurofibroma

Freckling in the a.'<illary or inguinal region

Optic pathway glioma

2 or more Lisch nodules (iris hamartomas)

Bony dysplasia, with or without bowing or pseudoarthrosis

First degree relative with NF1

Frequency13

>99%

>99

30-50%

85%

15%

90-95%

±3%

50%

Typical

age of onset

Congenital

> 7y

Congenital

>3y

<7y

>7y

Congenital

Not applicable

The phenotype of NFl is very variable, even within families. Although some patients only have

cqfo-au-lait macules, Lisch nodules and a few neurofibromas, other patients can display serious

complications. Frequent complications at pediatric age include disfigurement due to plexiform

neurofibromas, orthopedic problems (pseudoarthrosis (2%), scoliosis (10%)), endocrinologic

problems (5%; including precocious or delayed puberty and growth hormone deficiency),

cardiovascular problems (including pulmonary stenosis, and renal artery stenosis associated with

hypertension (2%)), and malignancy (including optic pathway gliomas (15%, see figure 1),

Juvenile myelomonocytic leukemia, low-grade central nervous system astrocytomas (2-3%),

malignant peripheral nerve sheet tumors (life-time risk 8-13%), and phaeochromocytoma

(2%)).13-19 The most common complication to affect quality of life in children with NFl,

however, are cognitive impairments,zo including mental retardation (4-8%), specific

neuropsychological deficits, learning disabilities and behavioral problems.21 The unpredictable

and diverse phenotype of NFl stresses the importance of age-specific monitoring by NF1

specialists.13

Genetic background

NFl is caused by a heterozygous mutation in the gene encoding for the neurofibromin protein

on chromosome 17q11.2.9, 1o The NF1 gene spans a region of about 335 kb of genomic DNA

and consists of over 60 exons. Several exons are alternatively spliced, including exon 9a and 23a.

The neurofibromin isoform with exon 9a is exclusively expressed in postmitotic forebrain

neurons,22 whereas the isoform containing exon 23a is expressed predominandy in glial cells.23

Figure 1. Clinical features of Neurofibromatosis type 1 (NFl).

Left: A toddler with familial NF1 with multiple caft·atr-lait maculae (dark arrow). Note the neurofibroma on the wrist of its

mother (white arrow). Right: Transversal T2 weighed MR (Magnetic Resonance) image of a glioma of the optic chiasm

(white arrow) in a 9-year old NFl patient.

The mutation rate of the NF1 gene is about 10-fold higher than that of other disease genes,

most probably because of its large size.24 The spectrum ofNF1 is very broad, with hundreds of

individual mutations identified so far, distributed over the different exons of the NF1 gene.

There are no clear mutational hotspots,24, 25 although several recurrent mutations and mutation

rich exons have been identified, together accounting for up to 30% of the mutations.2s About

half of all NF1 mutations result in premature termination codons,24 20-30% in splice defects,

and approximately 10% of the mutations are missense or single amino acid deletions.24. 25 About

5% of the NF1 patients have a microdeletion encompassing the entire NF1 gene and several

flanking genes, which is associated with a more severe cognitive and physical phenotype.26

N eurofibromin

The NF1 gene encodes for neurofibromin, a 2,818 amino acid protein which is expressed in a

wide array of cell types in the body, but is most abundant in neurons, Schwann cells and

CH.\PTER l 1 3

oligodendrocytes.27 Neurofibromin contains a GTP-ase activating protein (GAP) related

domain, which spans about 10% of the protein sequence.zs Through the GAP domain,

neurofibromin acts as a negative regulator of the activity of the RAS (rat sarcoma viral oncogene

homolog) proto-oncogenes. Thereby, neurofibromin functions as a tumor suppressor, which is

illustrated by the finding that benign and malignant tumor cell lines of NF1 patients exhibit a

decrease or loss of neurofibromin.29-33

By its action on RAS, neurofibromin downregulates the RAS/ERI< (Extracellular signal

regulated kinase) pathway34 and the RAS-PI3K (Phosphoinositide 3-kinase) / MTOR

(mammalian target of rapamycin) pathway,3S In addition, neurofibromin modulates the cAMP

(cyclic adenosine monophosphate) / PKA (cyclic AMP-dependent protein kinase A) pathway by

regulating Adenylyl Cyclase function in both RAS dependent36 and RAS independent ways.37-41

A simplified overview of the actions of neurofibromin is provided in figure 2.

~-

~ ~

(/) g ~ co s 0..

....... (]) .s: E

' + ..... ...._ <

I

•

Cell membrane

+ Adenyl ate Cyclase

~ ~

Proliferation- Differentiation

I I

• Figure 2: Simplified overview of the functions of neurofibromin. Evidence for action on the RAS pathway originates

predominantly from research in heterozygous Nj1 knockout mice and patient material, whereas evidence for the action on

the cAMP /PKA pathway originates mostly from research in homozygous Nf1 knockout drosophila.

A broad range of other properties and functions have been attributed to neurofibromin,

including an association with rnicrotubuli,42 a possible involvement in vesicle transport via its

interaction with Amyloid Precursor Protein,43 and a role in actin filament reorganisation,44

filopodia and dendritic spine formation4s, regulation of glial proliferation and neuronal

differentiation,3s and somatosensory cortex barrel formation.46 Like many other tumor

suppressors, neurofibromin is targeted to the nucleus.47 However, these functions are outside of

the scope of this thesis.

Relationship to other disorders

NF1 shows remarkable phenotypical overlap with other diseases grouped under the Neuro­

Cardio-Facial-Cutaneous (NCFC) syndromes and the Hamartoma syndromes, which include for

instance Noonan Syndrome, Costello Syndrome, and Tuberous Sclerosis Complex. These

diseases all involve some degree of cognitive impairment, cardiac defects, typical facial

dysmorphisms, macrocephaly, and cutaneous abnormalities. Most of these disorders are

associated with an increased risk of developing malignancies.4B, 49 The large overlap in clinical

phenotypes and the frequent lack of definite diagnostic criteria can make it difficult to establish

a diagnosis, especially in young children, when the disease phenotype is often not fully

developed. Recent advances in clinical generics have revealed the NCFC and Hamartoma

syndromes are all associated with mutations in genes in the RAS/ERK and RAS/PI3K/MTOR

pathways.4B, 49 Strikingly, despite the large overlap in generic background between these diseases,

even patients with identical mutations can display remarkably different phenotypes. 5° 51 Insights

into how the mutations found in NCFC and Hamartoma syndrome patients affect neuronal

signaling can facilitate the search for possible targeted treatments to alleviate the cognitive

burden of these syndromes.

Cognitive problems in NF1

Neuropsychological profile

The mean IQ of NFl patients is shifted to the left compared to the general population and

sibling controls, and ranges from the high SO's to the low 90's.zo, 52-60 As a result, NFl patients

have a two-fold increased risk at mental retardation (IQ below 70; 4-8%) compared to the

general population.Zl

In addition to a lower IQ, NFl is characterized by impairments across multiple

neuropsychological domains (thoroughly reviewed in sz, Gl-63). Deficits in visual spatial and visual

constructive skills, especially on the Judgment of Line Orientation Test, have long been

considered a hallmark of NFl.ZD, 58, 64-66 Other affected domains include executive functions

(such as planning and organization, and abstract concept formation), attention (divided,

switching and sustained), language (expressive and receptive) and memory (verbal, nonverbal

and tactile).ZD, 54, 56, 57, 67-71 However, problems with nonverbal and tactile memory could be

CII\P'!l,R 1 15

secondary to poor visual spatial skills or tactile perception, and not all reports confirm the

deficits in verbal memory.63 Several studies indicate that the problems in attention, visual spatial

skills and planning remain after correction for rq.zo. 57

Learning disabilities

Numerous studies have reported deficits in academic achievement tests administered in the

neuropsychological test setting, compared to normative scores or controls.s4, 56-58, 65-68, 72-74

Learning disabilities are reported in all academic areas, including reading, spelling and

mathematics (reviewed by Levine et al.62). Initial reports suggested the cognitive profile of NFl

resembles that of children with nonverbal learning disorder,sz which is characterized by

problems with motor behavior, social interaction, visual spatial skills and arithmetic, but not in

language.75 Although there are many parallels, this description does not quite fit. For instance,

literacy based learning disabilities turn out to occur at least as frequendy as mathematical

problems in NF1.S4. 67,73

Estimates for the prevalence of learning disabilities vary considerably (between 35 and 70%),21· 72

and suffer from small population sizes, selection bias, lack of control groups and differences in

the definitions for learning disability.s2 According to the DSM-IV criteria, a Specific Learning

Difficulty can be diagnosed only if academic achievement is more than two standart deviations

below the individual's level of intelligence (IQ).76 However, in NFl, low academic achievement

is frequendy seen in combination with a lower IQ.2o. 56,72 Thus, by only acknowledging learning

difficulties when patients show significant discrepancies between their academic achievement

and IQ, we would seriously underestimate the actual problems in learning experienced by NFl

children. 77

There is still little information about how NFl children function at school, where their cognitive

skills are put to the test in a setting that is in many ways different than the neuropsychological

test setting. Receiving intensive remedial teaching or special education may seriously confound

the interpretation of academic achievement test scores, and this may result in a significant

underestimation of the learning disabilities and school problems associated with NFL

Therefore, to get a more realistic assessment of school performance, it is important to combine

these different types of information on school functioning. However, quantitative studies on the

level of special education, remedial teaching, or grade repetition in these children are largely

absent.

1 6 lNTRODliCTION, AIMS eND OUTLINE

Behavior and social skills

NF1 patients commonly display difficulties in behavior and social skills. Parents and teachers

predominantly report behavioral problems in the internalizing domain (overcontrolled behavior

that contributes to distress to the child itself, such as anxiety, depression and withdrawal).S9. 67,78-

82 Some studies also reveal externalizing problems (undercontrolled behavior that contributes to

distress to others, such as aggressive behavior), albeit to a lesser extent than internalizing

problems.59, 78, 80-82 About 40% of the NFl patients meet diagnostic criteria for ADHD,2o. 78, 83

although only a small minority presents with the hyperactive subtype.20 Children with NFl and

ADHD were found to respond well to methylphenidate treatment, which improved attention,

behavior and social functioning.83 NFl has been associated with a higher incidence of autism

(about 4%),84 as well as other psychiatric and affective disorders, including dysthyrnia.85

Children with NFl are frequently reported to have poor social skills,78, 80. 81, 83 especially if they

have co-morbid ADHD_78 These difficulties are reflected in their interaction with other children,

as children with NFl are frequently picked on,59 have problems with peers,82 and have fewer

friends than other children.81 In addition, children with NFl are considered by teachers and

peers to be more sensitive and isolated, less likely to be leaders than other children,81 and to be

less independent than children without NF1.59, 86 They do, however, seem to display equal, or

even more pro-social behavior, such as being polite and helpful to others, compared to other

children. 81, 82

Interestingly, despite all problems mentioned above, children with NFl themselves report a

positive overall self concept,7B rate their own social skills as normal or above average,7B and have

an above average academic self-concept compared to objective norms.87 In addition, they do not

confirm the reports of teachers and peers on sensitivity and isolation, or leadership qualities.B1

Obviously, the effect of NFl on the experience of daily life is not straightforward. Barton et al.

proposed that children with NFl could have unrealistic positive self-perceptions, which

resembles the 'positive illusory bias' observed in otherwise healthy children with learning

disabilities or ADHD.87-B9 Possibly, this bias arises from a self-protective mechanism to prevent

confrontation with problems, or could reflect a focus on positive feedback only, or deficits in

processing feedback to one's own behavior.B7 In NFl, the latter could possibly be related to

NFl-specific neuropsychological dysfunction, such as their difficulty in interpreting social

cues,9o which could be secondary to problems in visual perceptual skills.

CH,\PTER 1 1 7

Quality of life

As can be expected from the numerous physical and cognitive problems associated with NFl,

adult patients report a below average quality of life (QOL) across multiple domains of skin

disease-specific and general health-related QOL questionnaires.91, 92 These problems include

emotional distress, inhibitions in physical and social functioning, and physical complaints. In

many domains these scores correlate to disease visibility, disease severity or both.91, 92 Parents of

toddlers with NF1 indicate that their children experience problems with growth and

development, physical functioning and behavior. In addition, parents themselves experience an

impact of their toddler's NFl on their personal time and emotions, and consider their child's

health to be below average.93 Another study revealed parents perceive problems in their child's

motor, cognitive, social and emotional functioning.79 The latter study is the only study so far

that has also investigated QOL self-reports of NF1 children, revealing that children experience

problems in the same areas as indicated by their parents, as well as in the domain of autonomy

(independence).79 One could imagine that the numerous cognitive and behavioral problems

have a substantial impact on QOL scores of NFl children. However, the relationship between

cognitive and behavioral problems, and QOL, has not been investigated yet.

Motor performance

Children with NF1 frequently display problems with :fine and gross motor functioning, including

problems with :fine motor speed, manual dexterity, balance and gait.2D, 53, 54, 56-58, 60, 68 Children

show a delay in reaching motor milestones,6S and are often described as being clumsy.17, 59 One

small study among 10 NFl patients suggested impairments in the latencies and directions of

saccadic eye movements.94

Unidentified Bright Objects

The most frequent NF1-related brain abnormalities are hyperintensities visible on T2 weighed

Magnetic Resonance (MR) or FLAIR (Fluid Attenuated Inversion Recovery) images, so-called

Unidentified Bright Objects (UBOs, see :figure 3). These UBOs are found in about 70% ofNFl

children, but tend to disappear in adulthood.61, 95 UBOs are predominantly found in the globus

pallidum, thalamus, cerebellum, brain stem and subcortical white matter.61 The differentiation

between UBOs and malignancies can be difficult. However, UBOs are not visible on CT or T1

weighed MR images, exert no mass effect, are not surrounded by edema, are not associated with

focal neurological deficits, and do not enhance with gadolineum contrast. Previous studies

showed that UBOs are not static and can disappear over time.96-9S

1 8 lNTROD\JCTION. "'I.!M:> \ND O!l'll~lNE

The exact nature of UBOs is unclear. One study performed a post-mortem histopathological

examination of brain areas that were diagnosed as UBOs on l'vllU, and reported spongiform

myelinopathy with vacuolar changes.99 These findings are confirmed by studies using Diffusion

Weighted Imaging. These studies report higher Apparent Diffusion Coefficients (ADC-values)

in NF1 brains at UBO-positive and UBO-negative sites compared to controls, suggesting an

increased water content in UBOs and normal appearing brain of NF1 patients.98,100-102 However,

the localization of this increased water content is still unclear.

Figure 3: Unidentified Bright Objects (UBOs).

Left: Transversal T2 weighed MR (Magnetic Resonance) image showing bilateral UBOs in the basal ganglia (arrows) in a 9-

year old NFl patient Right: Coronal FLAIR (Fluid Attenuated Inversion Recovery) image showing a large UBO in the left

cerebellar hemisphere (long arrow) and a smaller one in the right cerebellar hemisphere (short arrow) of a 2-year old NFl

patient.

Clinical correlates of cognitive problems in NF1

Genotype -phenotype relationships

It would be very useful if the cognitive abilities and disease severity of NF1 patients could be

predicted from their specific type of NF1 mutation. Unfortunately, studies aimed at finding

genotype-phenotype relationships are hampered by the broad mutational spectrum, and have

not been succesful.103-106 Only two strong relationships have been reported so far.

l'vficrodeletions of the NF1 gene are associated with a more severe clinical and cognitive

phenotype,107 and a 3bp in-frame deletion (c.2970-2972 delAAT) is reported to result in a

CH.iPTUZ 1 1 9

markedly mild clinical phenotype with possibly also a low frequency of learning disabilities. lOB

The more severe phenotype of microdeletion patients is possibly mediated by the deletion of

other genes in the microdeletion region. One of these, the RNF135 gene, was recently

postulated as a candidate gene for the overgrowth, facial dysmorphism and possibly the more

severe learning disabilities of NF1 microdeletic patients.109 The large variation in phenotypes,

even between individuals with identical mutations, suggests an important contribution of

modifier genes.JlO

Influence ofNFl-related brain abnonnalities on cognition

The presence of brain tumors in NF1 does not seem to be related to lower cognitive

functioning, unless children received radiotherapy, which has a strong negative impact.74 There

is no consensus on the relationship between UBOs and cognitive impairments in NFL Some

investigations reveal a connection between the presence, number, or localization of UBOs and

cognition,54, 56, 58, 60, 70, 111, 112 motor performance,6o, 111 and even between childhood UBOs and

adult cognitive performance.m However, others find no relationship_114, 115 Not all studies seem

sufficiently powered to justify conclusions. Importantly, heterozygous Nf1 mice show

impairments in learning and memory but do not display UBOs or other gross brain

abnormalities, at least not on a 4.7 Tesla MRI, indicating that the cognitive deficits in NF1 are

not necessarily related to gross anatomical changes.1. 116

Molecular and cellular mechanisms underlying cognitive deficits

Animal models have been of great help to delineate the molecular and cellular mechanisms

underlying cognitive deficits in NF1.

The etiology of cognitive deficits -lessons from Nfl mice

Nf1 heterozygous knockout mice display deficits in hippocampal-dependant learning and

memory, and attention.!, 7, 117 In addition, these mice show impaired hippocampal Long Term

Potentiation (LTP),1• ns which is an in vitro measure of synaptic plasticity, the process of

strengthening and weakening of neuronal contacts thought to be the neuronal substrate of

learning and memory_119 This deficit in LTP is observed when using a Theta Burst Stimulation

(TBS) protocol, but not when using a High Frequency stimulation protocol, which may hint to

increased sensitivity to GABA-agric inhibition. Indeed, GABA-agric inhibition was found to be

increased in Njt mice, and the GABA-A receptor antagonist picrotoxin can reverse the deficits

inLTP.1

2 0 TNTRODtiCTION, :\JMo .\ND 0UTI.INE

Importantly, the deficits in GABA-mediated inhibition, synaptic plasticity and learning in Nf1

mice can be rescued by genetically reducing the level of N-RAS or K-RAS, suggesting that the

cognitive phenotype of Nf1 mice is ultimately caused by enhanced RAS signaling.! Theoretically,

RAS activity can increase GABA (Gamma-aminobutyric acid)-mediated inhibition via

postsynaptic changes (for instance by regulating GABA-A receptor dynamics), or via

presynaptic changes (for instance by inducing the release of neurotransmitter vesicles containing

GABA). A presynaptic mechanism seems plausible in NF1, as studies using mutant mice

expressing the active H-R.as(G12V) gene revealed that active RAS facilitates neurotransmitter

release, through inducing ERI<:-mediated phosphorylation of Synapsin 1, a presynaptic protein

that is involved in neurotransmitter vesicle distribution.7 Therefore, the most plausible model is

that the cognitive phenotype of Nf1 mice results from increased RAS/ERI<:/Synapsin-I

signaling, leading to increased GABA release from inhibitory neurons, and impaired LTP (figure

4, left panel).

The Achilles' heel of RAS is that its activity is critically dependent upon its association to

membranes, for which it requires post-translational isoprenylation (i.e. addition of a farnesyl or

geranylgeranyl anchor).120 Thus, pharmacological reduction of RAS activity with Farnesyl

transferase inhibitors (FTI's) was found to restore the cognitive phenotype of Nj1 mice.! These

preclinical results offered perspectives at a drug therapy for cognitive impairments in humans, as

they indicated that the cognitive deficits in NF1 are due to reversible changes in synaptic

plasticity rather than structural anatomical abnormalities. However, since FTI's show significant

side effects, a drug needed to be found that had the same effect but was also safe enough for

long-term treatment of patients.

Statins decrease the synthesis of cholesterol, and isoprenoids (i.e. farnesyl and geranylgeranyl) by

inhibiting HMG-CoA reductase, the rate-limiting enzyme in the mevalonate synthesis pathway.

A breakthrough in the pursuit of a treatment for cognitive deficits in NF1 patients was made

when it was discovered that short-term treatment of Nj1 mice with lovastatin can reduce their

increased RAS activity, and thereby rescue their impairments in synaptic plasticity, learning and

memory, and attention7 (see figure 4, right panel).

Statins are used to treat hypercholesterolemia in millions of people worldwide, and have a

favorable safety profile in adults and children.121, 122 This preclinical proof of principle warranted

the initiation of translational clinical trials to assess the effect of statins on cognitive functioning

in NF1 patients, the first of which is described in chapter 7 of this thesis.

Cl-L\PTER 1 2 1

NF1

" tERK'\ ~

• [SYNIJ t

-4!iL­~~~

t Impaired

long- term potentiation

• Cognitive phenotype

Cell membrane

HMG-CoA

~ Farnesyl "/-. • •

• Cholesterol

NF1 + statins

• [SYNIJ t

Axon terminal ( ~ ~ • Restored

long-term potentiation

• Rescue of

cognitive phenotype

Figure 4: Simplified overview of the proposed presynaptic mechanism underlying the cognitive phenotype of Njl mice

and NFl patients, and the proposed mechanism through in which starins restore the cognitive phenotype of Njl mice.

Left: Heterozygous loss of neurofibromin function leads to elevated activity of RAS, which, via increased ERK-mediated

phosphorylation of synapsin 1, results in increased release of the inhibitory neurotransmitter GABA This increased

inhibition disturbs long-term potentiation, ultimately leading to the cognitive phenotype of NFl. Middle: Starins limit

HMG-CoA reductase, thereby reducing the production of isoprenoids including farnesyl. Right RAS activity is critically

dependent upon isoprenylation. Reducing farnesyl availability with statins in Njl mice normalizes the activity of RAS,

which, via reduced ERK-mediated phosphorylation of synapsin 1, results in normalization of the release of the inhibitory

neurotransmitter GABA. This restores long-term potentiation, ultimately rescuing the cognitive phenotype of Njl mice.

The etiology of cognitive deficits -lessons from drosophila

Interestingly, drosophila flies with a homozygous neurofibromin deletion show impairments in

immediate and long-term olfactory memory.4D, 123 The immediate memory deficit was shown to

be related to decreased cAMP signalling,4D whereas the long-term memory problems seems to be

related to a loss of inhibition of RAS.123 Strikingly, the cognitive phenotype of Nf1 drosophila,

like that ofNF1 mice, can also be rescued with statio treatment.124

Because immediate memory is not tested in the Nj1 mouse model the findings in drosophila

leave open the possibility that not all cognitive deficits in NF1 are due to enhanced RAS

2 2 TNTROD\ICTION, .\nr:; \ND 0tJTLINE

signaling. However, eDNA sequencing predicts drosophila neurofibromin has 60% amino acid

identity to human neurofibromin,39 whereas rnRNA sequencing predicts mouse neurofibromin

has over 98% amino acid conservation to human neurofibromin.125 In addition, Njt drosophila

studied have a homozygous loss of neurofibromin, whereas Njt mice still have one functional

copy of the NF1 gene. Possibly, this explains why not all changes observed in Nj1 flies can be

found in Njt mice.

The etiology of motor problems

A role for neurofibromin in motor functioning has been suggested by studies on mice with a

heterozygous deletion of exon 23a, which show impaired performance on a motor task (the

accelerating Rotarod test).126 Possibly, the motor problems in NF1 originate from the

cerebellum, as cerebellar Purkinje neurons are among the highest expressors of neurofibromin

in the brain, 23. 127 and the cerebellum is one of the predominant sites for UBOs,61 that have

been related to motor problems.6o. 111 Although NF1 patients are not clearly ataxic, their

clumsiness in movements could be related to deficits in the vermis, intermediate or lateral zones

of the cerebellum.J2B The cerebellum plays an important role in motor performance, but also in

motor learning, which refers to the ability to continuously adapt movements to optimize

performance, a task which requires neuronal plasticity.J29-13s The motor learning capacities of

children with NF1 have not been investigated so far.

CH\PTER 1 2 3

Aims

Children with NF1 display a wide variety of cognitive problems, including neuropsychological

deficits and problems with learning, behavior and motor performance. Preclinical studies

indicate these deficits are reversible, and have also identified statins as candidate drugs for the

treatment of cognitive deficits in NF1 patients.

The overall objectives of this thesis are to provide an overview of the impact of NF1 on daily

life, to identify possible outcome measures that can be used to assess potential therapeutic

illterventions, and to investigate the effect of statins on cognitive problems in NF1 patients.

These objectives are addressed in the following specific aims:

Aiml

To review the current knowledge of the etiology of cognitive deficits in NF1 and related

disorders within the Neuro-Cardio-Facial-Cutaneous and Hamartoma syndromes, and to review

potential treatment options.

Aim2

To provide insight into the impact of Neurofibromatosis type 1 on school performance.

Aim3

To assess parent- and child perceived Health Related Quality of Life in children with NF1, and

to identify potential targets for structural support.

Aim4

To examine motor problems in children with NF1, and to investigate whether these problems

arise from deficits in a specific brain area.

AimS

To explore the nature of T2-weighed hyperintensities on brain l'.1R imaging in NF1 patients.

Aim6

To assess the effect of simvastatin on neuropsychological, neurophysiological and

neuroradiological outcome measures in children with NF1

2 4 lNTRODLJCT!CJN, AIM,<; _,ND OUTLINE

Outline

The cognitive deficits of Neurofibromatosis type 1 (NF1) and several related disorders are

discussed in the light of the shared underlying molecular and cellular disturbances of these

diseases in chapter 2. The impact of NF1 on cognition and school performance, including need

for remedial teaching and special education is discussed in chapter 3. Child- and parent

perceived Quality of life, and potential determinants of reported problems are discussed in

chapter 4. Chapter 5 investigates whether impairments in motor functioning can be localized

to functional abnormalities in specific brain areas, whereas in chapter 6 focuses on the nature of

T2 weighed hyperintensities observed on brain MR. imaging in NF1 patients.

Chapter 7 reports the findings of a translational, randomized, double-blind, placebo-controlled

trial to investigate the effect of simvastatin on cognitive functioning in children with NF1, using

neuropsychological (chapter 3), neurophysiological (chapter 5) and neuroradiological (chapter

6) outcome measures.

Chapter 8 provides a discussion of the findings of this thesis, and a reflection on future

research prospectives. The results of this thesis are summarized in Chapter 9.

CH \PTTiR 1 2 5

References

1. Costa RM, Federov NB, Kogan JH, Murphy GG, Stern J, Ohno M, et al. Mechanism for the learning deficits in a

mouse model of neurofibromatosis type 1. Nature 2002;415:526-30.

2. Fernandez F, Morishita W, Zuniga E, Nguyen J, Blank M, Malenka RC, et al. Pharmacotherapy for cognitive

impairment in a mouse model of Down syndrome. Nat Neurosci 2007;10:411-3.

3. Guy J, Gan J, Selfridge J, Cobb S, Bird A. Reversal of neurological defects in a mouse model of Rett syndrome.

Science 2007;315:1143-7.

4. Yan QJ, Rammal M, Tranfaglia M, Bauchwitz RP. Suppression of two major Fragile X Syndrome mouse model

phenotypes by the mG!uRS antagonist MPEP. Neuropharmacology 2005;49:1053-66.

5. Ehninger D, HanS, Shilyansky C, Zhou Y, Li W, Kwiatkowski DJ, et al. Reversal oflearning deficits in a Tsc2(+ / -)

mouse model of tuberous sclerosis. Nat Med 2008.

6. Elgersma Y. Genetic engineering cures mice of neurological deficits: prospects for treating Angelman syndrome.

Pharmacogenomics 2007;8:539-41.

7. Li W, Cui Y, Kushner SA, Brown RA, Jentsch JD, Frankland PW, et al. The HMG-CoA reductase inhibitor

lovastatin reverses the learning and attention deficits in a mouse model of neurofibromatosis type 1. Curr Bioi

2005;15:1961-7.

8. Huson SM, Compston DA, Clark P, Harper PS. A genetic study of von Recklinghausen neurofibromatosis in south

east Wales. I. Prevalence, fitness, mutation rate, and effect of parental transmission on severity. J Med Genet

1989;26:704-11.

9. Viskochil D, Buchberg AM, Xu G, Cawthon RM, Stevens J, Wolff RK, et al. Deletions and a translocation interrupt

a cloned gene at the neurofibromatosis type 1locus. Cell1990;62:187-92.

10. Wallace MR., Andersen LB, Fountain]W, Odeh HM, Viskochil D, Marchuk DA, et al. A chromosome jump crosses

a translocation breakpoint in the von Recklinghausen neurofibromatosis region. Genes Chromosomes Cancer

1990;2:271-7.

11. National Institutes of Health Consensus Development Conference Statement: neurofibromatosis. Bethesda, Md.,

USA,July 13-15,1987. Neurofibromatosis 1988;1:172-8.

12. Cnossen MH, Moons KG, Garssen MP, Pasmans NM, de Goede-Bolder A, Niermeijer MF, et al. Minor disease

features in neurofibromatosis type 1 (NF1) and their possible value in diagnosis of NF1 in children < or = 6 years

and clinically suspected of having NFL Neurofibromatosis team of Sophia Children's Hospital. J Med Genet

1998;35:624-7.

13. Ferner RE, Huson SM, Thomas N, Moss C, Willshaw H, Evans DG, et al. Guidelines for the diagnosis and

management of individuals with neurofibromatosis 1. J Med Genet 2007 ;44:81-8.

14. Friedman JM. Neurofibromatosis 1: clinical manifestations and diagnostic criteria. J Child Neurol 2002;17:548-54;

discussion 71-2, 646-51.

15. Friedman JM, Arbiser J, Epstein JA, Gutmann DH, Huot SJ, Lin AE, et al. Cardiovascular disease in

neurofibromatosis 1: report of the NF1 Cardiovascular Task Force. Genet Med 2002;4:1 05-11.

16. Cnossen MH, Starn EN, Cooiman LC, Simonsz HJ, Stroink H, Oranje AP, et al. Endocrinologic disorders and optic

pathway gliomas in children with neurofibromatosis tJ:Pe 1. Pediatrics 1997;100:667-70.

17. Cnossen MH, de Goede-Bolder A, van den Broek KM, Waasdorp CM, Oranje AP, Stroink H, et al. A prospective 10

year follow up study of patients with neurofibromatosis type 1. Arch Dis Child 1998;78:408-12.

18. Evans DG, Baser ME, McGaughran J, Sharif S, Howard E, Moran A. Malignant peripheral nerve sheath tumours in

neurofibromatosis 1. J Med Genet 2002;39:311-4.

2 6 TNTROD!.'CTTON. i\JMS .\ND 0\JTJJNE

19. Drouet A, Wolkenstein P, Lefaucheur JP, Pinson S, Combemale P, Gherardi RK, et al. Neurofibromatosis 1-

associated neuropathies: a reappraisal Brain 2004;127:1993-2009.

20. Hyman SL, Shores A, North KN. The narure and frequency of cognitive deficits in children with neurofibromatosis

type 1. Neurology 2005;65:1037 -44.

21. North KN, Riccardi V, Samango-Sprouse C, Ferner R, Moore B, Legius E, et a!. Cognitive function and academic

performance in neurofibromatosis. 1: consensus statement from the NF1 Cognitive Disorders Task Force.

Neurology 1997;48:1121-7.

22. Gutmann DH, Zhang Y, Hirbe A. Developmental regulation of a neuron-specific neurofibromatosis 1 isoform. Ann

Neurol1999;46:777-82.

23. Gutmann DH, Geist RT, Wright DE, Snider WD. Expression of the neurofibromatosis 1 (NF1) isoforms in

developing and adult rat tissues. Cell Growth Differ 1995;6:315-23.

24. Fahsold R, Hoffmeyer S, M:ischung C, Gille C, Ehlers C, Kucukceylan N, et al. Minor lesion mutational spectrum of

the entire NF1 gene does not explain its high mutability but points to a functional domain upstream of the GAP­

related domain. AmJ Hum Genet 2000;66:790-818.

25. Messiaen LM, Callens T, Mortier G, Beysen D, Vandenbroucke I, Van Roy N, et al. Exhaustive mutation analysis of

the NF1 gene allows identification of 95% of mutations and reveals a high frequency of unusual splicing defects.

Hum Mutat 2000;15:541-55.

26. Kluwe L, Siebert R, Gesk S, Friedrich RE, Tinschert S, Kehrer-Sawatzki H, et al. Screening 500 unselected

neurofibromatosis 1 patients for deletions of the NF1 gene. Hum Mutat 2004;23:111-6.

27. Daston MM, Scrable H, Nordlund M, Sturbaum AK, Nissen LM, Ratner N. The protein product of the

neurofibromatosis type 1 gene is expressed at highest abundance in neurons, Schwann cells, and oligodendrocytes.

Neuron 1992;8:415-28.

28. Martin GA, Viskochil D, Bollag G, McCabe PC, Crosier WJ, Haubruck H, et al. The GAP-related domain of the

neurofibromatosis type 1 gene product interacts with ras p21. Cell1990;63:843-9.

29. Legius E, Marchuk DA, Collins FS, Glover TW. Somatic deletion of the neurofibromatosis type gene in a

neurofibrosarcoma supports a tumour suppressor gene hypothesis. Nat Genet 1993;3:122-6.

30. DeClue JE, Papageorge AG, Fletcher JA, Diehl SR, Ratner N, Vass WC, et al. Abnormal regulation of mammalian

p21ras contributes to malignant tumor growth in von Recklinghausen (type 1) neurofibromatosis. Cell 1992;69:265-

73.

31. Basu TN, Gutmann DH, Fletcher JA, Glover TW, Collins FS, Downward]. Aberrant regulation of ras proteins in

malignant tumour cells from type 1 neurofibromatosis patients. Nature 1992;356:713-5.

32. Gutmann DH, Donahoe J, Brown T, James CD, Petty A. Loss of neurofibromatosis 1 (NF1) gene expression in

NF1-associated pilocytic astrocytomas. Neuropathol Appl Neurobiol2000;26:361-7.

33. Maertens 0, Brems H, Vandesompele J, De Raedt T, Heyns I, Rosenbaum T, et al. Comprehensive NF1 screening

on cultured Schwann cells from neurofibromas. Hum Murat 2006;27:1030-40.

34. Dasgupta B, Gutmann DH. Neurofibromatosis 1: dosing the GAP between mice and men. Curr Opin Genet Dev

2003;13:20-7.

35. Johannessen CM, Reczek EE, James MF, Brems H, Legius E, Cichowski K The NF1 tumor suppressor critically

regulates TSC2 and mTOR. Proc Nat! Acad Sci U SA 2005;102:8573-8.

36. Hannan F, HoI, Tong JJ, Zhu Y, Numberg P, Zhong Y. Effect of neurofibromatosis type I mutations on a novel

pathway for adenylyl cyclase activation requiting neurofibromin and Ras. Hum Mol Genet 2006;15:1087-98.

37. Guo HF, The I, Hannan F, Bernards A, Zhong Y. Requirement of Drosophila NF1 for activation of adenylyl cyclase

by PACAP38-like neuropeptides. Science 1997;276:795-8.

C:H.\PTER I 2 7

38. Hegedus B, Dasgupta B, Shin JE, Emnett RJ, Hart-Mahon EK, Elghazi L, et al. Neurofibromatosis-! regulates

neuronal and glial cell differentiation from neuroglial progenitors in vivo by both c.ANIP- and Ras-dependent

mechanisms. Cell Stem Cell2007;1:443-57.

39. The I, Hannigan GE, Cowley GS, Reginald S, Zhong Y, Gusella JF, et al. Rescue of a Drosophila NFl mutant

phenotype by protein kinase A. Science 1997;276:791-4.

40. Guo HF, Tong J, Hannan F, Luo L, Zhong Y. A neurofibromatosis-1-regulated pathway is required for learning in

Drosophila. Nature 2000;403:895-8.

41. Tong J, Hannan F, Zhu Y, Bernards A, Zhong Y. Neurofibromin regulates G protein-stimulated adenylyl cyclase

activity. Nat Neurosci 2002;5:95-6.

42. Xu H, Gutmann DH. Mutations in the GAP-related domain impair the ability of neurofibromin to associate with

microtubules. Brain Res 1997;759:149-52.

43. De Schepper S, Boucneau JM, Westbroek W, Mommaas M, Onderwater J, Messiaen L, et al. Neurofibromatosis type

1 protein and amyloid precursor protein interact in normal human melanocytes and colocalize with melanosomes. J

Invest Dermatol2006;126:653-9.

44. Ozawa T, Araki N, Yunoue S, Tokuo H, Feng L, Patrakitkomjom S, et al. The neurofibromatosis type 1 gene

product neurofibromin enhances cell motility by regulating actin filament dynamics via the Rho-ROCK-LIMK2-

cofilin pathway. J Bioi Chern 2005;280:39524-33.

45. Lin YL, Lei YT, Hong CJ, Hsueh YP. Syndecan-2 induces filopodia and dendritic spine formation via the

neurofibromin-PKA-Ena/VASP pathway. J Cell Biol2007;177:829-41.

46. Lush l\IIE, Li Y, Kwon CH, Chen J, Parada LF. Neurofibromin is required for barrel formation in the mouse

somatosensory cortex. J Neurosci 2008;28:1580-7.

47. Vandenbroucke I, Van Oostveldt P, Coene E, De Paepe A, Messiaen L. Neurofibromin is actively transported to the

nucleus. FEBS Lett 2004;560:98-102.

48. Bentires-Alj M, Kontaridis MI, Nee! BG. Stops along the RAS pathway in human genetic disease. Nat Med

2006;12:283-5.

49. Denayer E, Legius E. What's new in the neuro-cardio-facial-cutaneous syndromes? Eur J Pediatr 2007;166:1091-8.

50. Nystrom AM, Ekvall S, Berglund E, Bjorkqvist M, Braathen G, Duchen K, et al. Noonan and cardio-facial­

cutaneous syndromes: two clinically and genetically overlapping clisorders. J Med Genet 2008;45:500-6.

51. Carta C, Pantaleoni F, Bocchinfuso G, Stella L, Vasta I, Sarkozy A, et al. Germline missense mutations affecting

KRAS Isoform Bare associated with a severe Noonan syndrome phenotype. Am J Hum Genet 2006;79:129-35.

52. Ozonoff S. Cognitive impairment in neurofibromatosis type 1. Am J Med Genet 1999;89:45-52.

53. Eldridge R, Denckla MB, Bien E, Myers S, Kaiser-Kupfer MI, Pikus A, et al Neurofibromatosis type 1

(Recklinghausen's disease). Neurologic and cognitive assessment with sibling controls. Am J Dis Child 1989;143:833-

7.

54. Hofman KJ, Harris EL, Bryan RN, Denckla MB. Neurofibromatosis type 1: the cognitive phenotype. J Pediatr

1994;124:Sl-8.

55. Legius E, Descheemaeker MJ, Spaepen A, Casaer P, Fryns JP. Neurofibromatosis type 1 in childhood: a study of the

neuropsychological profile in 45 children. Genet Couns 1994;5:51-60.

56. North K, Joy P, Yuille D, Cocks N, Mobbs E, Hutchins P, et al. Specific learning disability in children with

neurofibromatosis type 1: significance ofMRI abnormalities. Neurology 1994;44:878-83.

57. Ferner RE, Hughes RA, Weinman]. Intellectual impairment in neurofibromatosis 1. J Neurol Sci 1996;138:125-33.

58. Moore BD, Slopis JM, Schomer D, Jackson EF, Levy BM. Neuropsychological significance of areas of high signal

intensity on brain MRis of children with neurofibromatosis. Neurology 1996;46:1660-8.

2 8 INTRODUCTlON. c\l~fO' _\ND Ol.'TUNF

59. Dilts CV, Carey JC, Kircher JC, Hoffman RO, Creel D, Ward K, et al. Children and adolescents with

new:ofibromatosis 1: a behavioral phenotype. J Dev Behav Pediatr 1996;17:229-39.

60. Feldmann R, Denecke J, Grenzebach M, Schuierer G, Weglage J. New:ofibromatosis type 1: motor and cognitive

function and T2-weighted MRl hyperintensities. New:ology 2003;61:1725-8.

61. North K. New:ofibromatosis type 1. AmJ Med Genet 2000;97:119-27.

62. Levine TM, Materek A, Abel J, O'Donnell M, Cutting LE. Cognitive profile of new:ofibromatosis type 1. Semin

Pediatr New:ol 2006;13:8-20.

63. North K, Hyman S, Barton B. Cognitive deficits in new:ofibromatosis 1. J Child New:ol 2002;17:605-12; discussion

27-9, 46-51.

64. Schrimsher GW, Billingsley RL, Slopis JM, Moore BD, 3rd. Visual-spatial performance deficits in children with

new:ofibromatosis type-1. AmJ Med GenerA 2003;120A:326-30.

65. Billingsley RL, Jackson EF, Slopis JM, Swank PR, Mahankali S, Moore BD. Functional MRl of visual-spatial

processing in new:ofibromatosis, type I. New:opsychologia 2004;42:395-404.

66. Clements-Stephens AM, Rimrodt SL, Gaw: P, Cutting LE. Visuospatial processing in children with

new:ofibromatosis type 1. New:opsychologia 2008;46:690-7.

67. Descheemaeker MJ, Ghesquiere P, Symons H, Fryns JP, Legius E. Behavioural, academic and new:opsychological

profile of normally gifted New:ofibromatosis type 1 children. J Intellect Disabil Res 2005;49:33-46.

68. Chapman CA, Waber DP, Bassett N, Urion DK, Korf BR. New:obehavioral profiles of children with

new:ofibromatosis 1 referred for learning disabilities are sex-specific. AmJ Med Genet 1996;67:127-32.

69. McCaffrey RJ, Westervelt HJ. Issues associated with repeated new:opsychological assessments. New:opsychol Rev

1995;5:203-21.

70. Denckla MB, Hofman I(, Mazzocco MM, Melhem E, Reiss AL, Bryan RN, eta!. Relationship between T2-weighted

hyperintensities (unidentified bright objects) and lower IQs in children with new:ofibromatosis-1. AmJ Med Genet

1996;67:98-102.

71. Varnhagen CI(, Lewin S, Das JP, Bowen P, Ma I(, Klimek M. New:ofibromatosis and psychological processes. J Dev

Behav Pediatr 1988;9:257 -65.

72. Brewer VR, Moore BD, 3rd, Hiscock M. Learning disability subtypes in children with new:ofibromatosis. J Learn

Disabil1997;30:521-33.

73. Hyman SL, Atthw: Shores E, North KN. Learning disabilities in children with new:ofibromatosis type 1: subtypes,

cognitive profile, and attention-deficit-hyperactivity disorder. Dev Med Child New:ol 2006;48:973-7.

7 4. De Winter AE, Moore BD, 3rd, Slopis JM, Ater JL, Copeland DR. Brain tumors in children with new:ofibromatosis:

additional new:opsychological morbidity? New:o Oncol1999;1:275-81.

75. Harnadek MC, Row:ke BP. Principal identifying featw:es of the syndrome of nonverbal learning disabilities in

children. J Learn Disabi11994;27:144-54.

76. American Psychiatric Association. Diagnostic and statistical manual of mental disorderss: DSM-IV. 4 ed.

Washington, DC: American Psychiatric Association; 1994.

77. Watt SE, Shores EA, North KN. An Examination of Lexical and Sublexical Reading Skills in Children with

New:ofibromatosis Type 1. Child New:opsychol2007:1-18.

78. Barton B, North K. Social skills of children with new:ofibromatosis type 1. Dev Med Child New:ol2004;46:553-63.

79. Graf A, Landolt MA, Mori AC, Boltshauser E. Quality of life and psychological adjustment in children and

adolescents with new:ofibromatosis type l.J Pediatr 2006;149:348-53.

80. Johnson NS, Saal HM, Lovell AM, Schorry EK. Social and emotional problems in children with new:ofibromatosis

type 1: evidence and proposed interventions. J Pediatr 1999;134:7 67-72.

CH. \PTER 1 2 9

81. Noll RB, Reiter-Purtill J, Moore BD, Schorry EK, Lovell AM, Vannatta K, et al. Social, emotional, and behavioral

functioning of children with NFL Am] Med Genet A 2007;143A:2261-73.

82. Johnson H, Wiggs L, Stores G, Huson SM. Psychological disturbance and sleep disorders in children with

neurofibromatosis type 1. Dev Mcd Child Neuro\2005;47:237-42.

83. Maumer VF, I<.luwe L, Thakker SD, Leark RA. Treatment of ADHD in neurofibromatosis type 1. Dev Med Child

Neuro\2002;44:164-70.

84. Williams PG, Hersh JH. Brief report: the association of neurofibromatosis type 1 and autism. J Autism Dev Disord

1998;28:567-71.

85. Zoller ME, Rembeck B. A psychiatric 12-year follow-up of adult patients with neurofibromatosis type 1. J Psychiatr

Res 1999;33:63-8.

86. Prinzie P, Descheemaeker MJ, Vogels A, Cleymans T, Haselager GJ, Curfs LM, et al. Personality profiles of children

and adolescents with neurofibromatosis type 1. Am] Med Genet A 2003;118A:1-7.

87. Barton B, Norrh K The self-concept of children and adolescents with neurofibromatosis type 1. Child Care Health

Dev 2007;33:401-8.

88. Owens JS, Hoza B. The role of inattention and hyperactivity/impulsivity in the positive illusory bias. J Consult Clin

Psycho\2003;71:680-91.

89. Heath NL, Glen T. Positive illusory bias and the self-protective hypothesis in children with learning disabilities. J

Clin Child Adolesc Psycho! 2005;34:272-81.

90. Eliason MJ. Neurofibromatosis: implications for learning and behavior. J Dev Behav Pediatr 1986;7:175-9.

91. Page PZ, Page GP, Ecosse E, Korf BR, Leplege A, Wolkenstein P. Impact of neurofibromatosis 1 on Quality of

Life: a cross-sectional study of 176 American cases. Am J Med Genet A 2006;140:1893-8.

92. Wolkenstein P, Zeller J, Revu.z J, Ecosse E, Leplege A. Quality-of-life impairment in neurofibromatosis type 1: a

cross-sectional study of128 cases. Arch Dermato\2001;137:1421-5.

93. Oostenbrink R, Spong K, de Goede-Bolder A, Landgraf JM, Raat H, Moll HA. Parental reports of health-related

quality of life in young children with neurofibromatosis type 1: influence of condition specific determinants. J Pediarr

2007;151:182-6, 6 e1-2.

94. Lasker AG, Denckla MB, Zee DS. Ocular motor behavior of children with neurofibromatosis 1. J Child Neurol

2003;18:348-55.

95. Gill DS, Hyman SL, Steinberg A, Norrh KN. Age-related findings on MRI in neurofibromatosis type 1. Pediatr

Radio\2006;36:1048-56.

96. Kraut MA, Gerring JP, Cooper KL, Thompson RE, Denckla MB, Kaufmann WE. Longitudinal evolution of

unidentified bright objects in children with neurofibromatosis-1. Am J Med Genet A 2004;129:113-9.

97. Imamura A, Matsuo N, Okuda M, Morira H, Iwata M, Yamazaki Y, et al. Serial MR imaging and (1)H-MR

spectroscopy of unidentified bright objects in a case of neurofibromatosis type 1. Brain Dev 2005.

98. Tognini G, Ferrozzi F, Garlaschi G, Piazza P, Patti A, Virdis R, et al. Brain apparent diffusion coefficient evaluation

in pediatric patients with neurofibromatosis type 1. J Comput Assist Tomogr 2005;29:298-304.

99. DiPaolo DP, Zimmerman RA, Rorke LB, Zackai EH, Bilaniuk LT, Yachnis AT. Neurofibromatosis type 1:

pathologic substrate of high-signal-intensity foci in the brain. Radiology 1995;195:721-4.

100. Alkan A, Sigirci A, Kutlu R, Ozcan H, Erdem G, Asian M, et al. Neurofibromatosis type 1: diffusion weighted

imaging findings of brain. EurJ Radio\2005;56:229-34.

101. Sheikh SF, Kubal WS, Anderson AW, Muralik P. Longitudinal evaluation of apparent diffusion coefficient in

children with neurofibromatosis type 1. J Comput Assist Tomogr 2003;27:681-6.

102. Eastwood JD, Fiorella DJ, MacFall JF, Delong DM, Provenzale JM, Greenwood RS. Increased brain apparent

diffusion coefficient in children with neurofibromatosis type 1. Radiology 2001;219:354-8.

3 0 lNTR(JDIY:110N. ,\JMS _\ND OliTLINl'

103. Cnossen MH, van der Est MN, Brenning MH, van Asperen CJ, Breslau-Siderius EJ, van der Ploeg AT, et al.

Deletions spanning the neurofibromatosis type 1 gene: implications for genotype-phenotype correlations in

neurofibromatosis type 1? Hum Mutat 1997;9:458-64.

104. Wimmer K, Muhlbauer M, Eckart M, Callens T, Rehder H, Birkner T, et al. A patient severely affected by spinal

neurofibromas carries a recurrent splice site mutation in the NFl gene. Eur J Hum Genet 2002;10:334-8.

105. Szudek J, Joe H, Friedman JM. Analysis of intrafamilial phenotypic variation in neurofibromatosis 1 (NF1). Genet

Epidemiol2002;23:150-64.

106. Kaufmann D, Muller R, Bartelt B, Wolf M, Kunzi-Rapp K, Hanemann CO, et al Spinal neurofibromatosis without

cafe-au-lait macules in two families with null mutations of the NF1 gene. Am J Hum Genet 2001;69:1395-400.

107. Descheemaeker MJ, Roelandts K, De Raedt T, Brems H, Fryns JP, Legius E. Intelligence in individuals with a

neurofibromatosis type 1 microdeletion. Am] Med Genet A 2004;131:325-6.

108. Upadhyaya M, Huson SM, Davies M, Thomas N, Chuzhanova N, Giovannini S, et al. An absence of cutaneous

neurofibromas associated with a 3-bp inframe deletion in exon 17 of the NF1 gene (c.2970-2972 delAAT): evidence

of a clinically significant NF1 genotype-phenotype correlation. Am] Hum Genet 2007;80:140-51.

109. Douglas J, Cilliers D, Coleman K, Tatron-Brown K, Barker K, Bernhard B, et al. Mutations in RNF135, a gene

within the NF1 microdeletion region, cause phenotypic abnormalities including overgrowth. Nat Genet 2007;39:963-

5.

110. Easton DF, Ponder JVIA, Huson SM, Ponder BA. An analysis of variation in elcl'ression of neurofibromatosis (NF)

type 1 (NF1): evidence for modifying genes. Am] Hum Genet 1993;53:305-13.

111. Goh WH, Khong PL, Leung CS, Wong VC. T2-weighted hyperintensities (unidentified bright objects) in cbildren

with neurofibromatosis 1: their impact on cognitive function. J Child Neurol2004;19:853-8.

112. Hyman SL, Gill DS, Shores EA, Steinberg A, Norrh KN. T2 hyperintensities in children with neurofibromatosis type

1 and their relationship to cognitive functioning. J Neural Neurosurg Psychiatry 2007;78:1088-91.

113. Hyman SL, Gill DS, Shores EA, Steinberg A, Joy P, Gibikote SV, et al Natural history of cognitive deficits and their

relationship to M:R.I T2-hyperintensities in NFL Neurology 2003;60:1139-45.

114. Ferner RE, Chaudhuri R, Bingham], Cox T, Hughes RA. M:R.I in neurofibromatosis 1. The nature and evolution of

increased intensity T2 weighted lesions and their relationship to intellecrual impairment. J Neurol Neurosurg

Psychistty 1993;56:492-5.

115. Legius E, Descheemaeker MJ, Steyaert J, Spaepen A, Vlietinck R, Casaer P, et al. Neurofibromatosis type 1 in

childhood: correlation ofM:R.I findings with intelligence. J Neurol Neurosurg Psychiatry 1995;59:638-40.

116. Rosenbaum T, Engelbrecht V, Krolls W, van Dorsten FA, Hoehn-Berlage M, Lenard HG. M:R.I abnormalities in

neurofibromatosis type 1 (NF1): a study of men and mice. Brain Dev 1999;21:268-73.

117. Silva AJ, Frankland PW, Marowitz Z, Friedman E, Laszlo GS, Cioffi D, et al. A mouse model for the learning and

memory deficits associated with neurofibromatosis type I. Nat Genet 1997;15:281-4.

118. Guilding C, McNair K, Stone TW, Morris BJ. Restored plasticity in a mouse model of neurofibromatosis type 1 via

inhibition of hyperactive ERK and CREB. Eur J Neurosci 2007;25:99-105.

119. Kandel ER. The molecular biology of memory storage: a dialogue between genes and synapses. Science

2001;294:1030-8.

120. Konstantinopoulos PA, Karamouzis MV, Papavassiliou AG. Post-translational modifications and regulation of the

RAS superfamily of GTPases as anticancer targets. Nat Rev Drug Discov 2007;6:541-55.

121. de Jongh S, Ose L, Szamosi T, Gague C, Lambert M, Scott R, et al. Efficacy and safety of statin therapy in children

with familial hypercholesterolemia: a randomized, double-blind, placebo-controlled trial with simvastatin. Circulation

2002;106:2231-7.

122. Menge T, Hartung HP, Smve 0. Statins--a cure-all for the brain? Nat Rev Neurosci 2005;6:325-31.

0-f\PTER 1 3 1

123. Ho IS, Hannan F, Guo HF, Bakker I, Zhong Y. Distinct functional domains of neurofibromatosis type 1 regulate

immediate versus long-term memory formation. J Neurosci 2007;27:6852-7.

124. Vose LR, Arora GS, O'Brien PF, Hannan F. Pharmacologic Rescue of Behavioral Deficits in Drosophila NF1

Mutants. NF conference (abstract) 2008 p97.

125. Bernards A, Snijders AJ, Hannigan GE, Murthy AE, Gusella JF. Mouse neurofibromatosis type 1 eDNA sequence

reveals high degree of conservation of both coding and non-coding mRNA segments. Hum Mol Genet 1993;2:645-

50.

126. Costa RM, Yang T, Huynh DP, Pulst SM, Viskochil DH, Silva AJ, et al. Learning deficits, but normal development

and tumor predisposition, in mice lacking exon 23a ofNf1. Nat Genet 2001;27:399-405.

127. Nordlund ML, Rizvi TA, Brannan CI, Ratner N. Neurofibromin expression and astrogliosis in neurofibromatosis

(type 1) brains. J Neuropathol Exp Neurol1995;54:588-600.

128. Gazzaniga MS, Ivry RB, Mangun GR. Cognitive Neuroscience- The biology of the mind, Chapter 10. New York­

London: W.W. Norton & Company; 1998.

129. Koekkoek SK, Hulscber HC, Dortland BR, Hensbroek RA, Elgersma Y, Ruigrok TJ, et a!. Cerebellar LTD and

learning-dependent timing of conditioned eyelid responses. Science 2003;301:1736-9.

130. Hopp JJ, Fuchs AF. The characteristics and neuronal substrate of saccadic eye movement plasticity. Prog Neurobiol

2004;72:27 -53.

131. Barash S, Mehl:yan A, Sivakov A, Zhang M, Glickstein M, Thier P. Saccadic dysmetria and adaptation after lesions of

the cerebellar cortex.] Neurosci 1999;19:10931-9.

132. Desmurget M, Pelisson D, Grethe JS, Alexander GE, Urquizar C, Prablanc C, et al. Functional adaptation of reactive

saccades in humans: a PET study. Exp Brain Res 2000;132:243-59.

133. Baizer JS, Kralj-Hans I, Glickstein M. Cerebellar lesions and prism adaptation in macaque monkeys. J Neurophysiol

1999;81:1960-5.

134. Pis ella L, Rossetti Y, Michel C, Rode G, Boisson D, Pelisson D, eta!. Ipsidirectional impairment of prism adaptation

after unilateral lesion of anterior cerebellum. Neurology 2005;65:150-2.

135. Clower DM, Hoffman JM, Votaw JR, Faber TL, Woods RP, Alexander GE. Role of posterior parietal cortex in the

recalibration of visually guided reaching. Nature 1996;383:618-21.

3 2 INTRODUCTION, AlMS ,\ND 0\JHJNE

Lianne C. Ktab1.2*, SusannaM.I. Goorden1*, YpeElgersma1

*Authors contributed equally

1 Departme11t of New:oscie11ce; 2 Departme11t ofGe11eral Pediatrics, Erasmus MC Umversity Medical Ce11ter- Sophia Cllildre11's Hospital, The Netherla11ds

Trends in Genetics 2008;24(10): 498-510

Summary

Defects in rat sarcoma viral oncogene homolog (RAS)-extracellular signal regulated kinase

(ERK) and phosphati.dylinositol 3-kinase (PI3K)-mammalian target of rapamycin (MTOR)

signaling pathways have recently been shown to cause several genetic disorders classified as

neuro-cardio-facial-cutaneous (NCFC) and Hamartoma syndromes. Although these pathways

are well-known players in cell proliferation and cancer, their role in cognitive function is less

appreciated. Here, we focus on the cognitive problems associated with mutations in the RAS­

ERK. and PI3K-MTOR signaling pathways and on the underlying mechanisms revealed by

recent animal studies. Cancer drugs have been shown to reverse the cognitive deficits in mouse

models ofNCFC and Hamartoma syndromes, raising hopes for clinical trials.

3 4 ONCOGENES ON MY MIND: ERK AND MTOR SIGNAl.JNG IN COGNITIVE DISEA.SES

Mutations in RAS signaling pathways are a leading cause for cognitive dysfunction

The RAS (rat sarcoma viral oncogene homolog) signaling pathways are evolutionary conserved

pathways, transducing signals from membrane-bound receptors to proteins that regulate

fundamental cell processes like cell growth and proliferation. Therefore, it is not surprising that

genetic disorders with gain-of-function mutations in the RAS signaling pathways are

characterized by benign and malignant overgrowths. This is a common phenotype for two

groups of syndromes, classified in neuro-cardio-facial-cutaneous (NCFC) and Hamartoma

syndromes. A high prevalence of mental retardation (see Glossary) and behavioral disturbances

is also found among these patients (Table 1). Many of the genes associated with these diseases

have been identified in the past few years and all are part of the ERK (extracellular signal

regulated kinase) and MTOR (mammalian target of rapamycin) pathways (Figure 1 at the end of

the chapter). For example, mutations in 5051 (Son of Sevenless, Drosophila, homolog 1) result in

Noonan Syndrome, whereas mutations in RAF-1 (v-raf-1 murine leukemia viral oncogene

homolog 1) result in Noonan syndrome and LEOPARD, and mutations in 5PRED-1 (sprouty

related EVH1 domain containing protein 1) in Neurofibromatosis type I-like syndrome.l-11

Genetic alterations in RAS-ERI<. and PI3K (phosphoionositide 3-kinase)-MTOR signaling can

be considered a leading cause of cognitive and behavioral impairments, collectively affecting ~

1/1000 people.

Studies on rodents carrying mutations in components of the RAS-ERK signaling pathways

indicate that postmitotic neurons have reprogrammed these signaling pathways to regulate

synaptic plasticity (fable 2), believed to be the cellular basis for learning and memory (Figure 2

at the end of the chapter). Combining these neuroscience studies with molecular insights from

cancer research has rapidly increased our understanding of the etiology of the cognitive deficits

in the affected patients, and offers the opportunity to treat the cognitive deficits in NCFC and

Hamartoma syndrome patients.

Cognitive deficits arising from genetic impairments in RAS-ERK signaling

The NCFC syndromes comprise a constellation of disorders that include Neurofibromatosis

Type 1 (NF1), Noonan syndrome, Costello syndrome, Cardia-Facial-Cutaneous (CFC)

syndrome, LEOPARD syndrome (an acronym for its cardinal features; lentigines, ECG

conduction abnormalities, ocular hypertelorism, pulmonic stenosis, abnormal genitalia,

retardation of growth, and sensorineural deafness) and NF1-Like syndrome. All these

syndromes are associated with some degree of mental impairment (Table 1). In general,

C!! \Pl1cR 2 3 5

(.N Table 1: Cognitive phenotypes ofNCFC and Hamartoma syndromes•

0'\ Disease (prevalence) Prominent Genes e1o of cases Protein (function) CNS features• phenotypical associated with characteristics gene)b Very frequent Frequent Less frequent

0 (75-100%) (25-74%) (up to25%)

z Neuro-cardlo-facial-cutaneous syndromes n

0 C'l Neurofibromatosis type Cafl:-au-lait macules, NFI (95%)" Neurofibromin (ltAS-GAP) ' Low-average IQ," Learning disabilities, ADHD 1\Iild tv!R, autism, tli z 1 skin fold freckling, specific deficits in a, social, emotional and seizures, low grade tli (f> (NFl, 1:3000) Lisch nodules, attention, executive behavioral problems, motor gliomas13, 47,49 0 cutaneous and functioning and problems, speech problems, z "' plexiform visual-spatial skills 13 sleep disturbances, ~1RI a

::< neurofibromas52 abnormalities, i!:: macroccphaly13,4? z

Neurofibromatosis 1- Cafl:-au-lait macules, SPREDI" SPREDl (Inhibitor ofRaf Frequency !=? til like syndrome (rare) skin-fold freckling, activation by RAS) unknown. Some

~ macrocephaly, lipomas; patients with

~ no neurofibromas or macrocephaly, Lisch nodules! I learning disabilities

1:1

is: Noonan syndrome Typical facial features PIPN/t (50%)' SHP2 (tyrosine phosphatase) Low-average IQ12 Learning disabilities, motor liiild MR, social >-! (1:2000) (hypertclorism, ptosis, RAFt (3-17%)~' RAFl (setine/threoninc kinase) problems, speech and etnotional 0 low-set posteriorly BRAF(<2%)' BRAF (serine/ threonine kinase) problems69 problems, ::e w rotated ears), webbed KRAS (-2%)4,10,62 KRAS (small G-protein) seizuresl2. 69,7°

~ neck, short stature, SOS/ (-9-13%)'·7 SOSl (GEF protein)'

~ cardiac probletns 52 J.fEKI (-<2%)6' MEKl (tyrosine/ setine/

threonine kinase)

C'l LEOPARD Multiple lentigines, PIPN/1 (>80%)7' SHP2 (tyrosine phosphatase) tvWdl\msz

z (rare)< cardiac problems, short RAFt (-<7%)' RAFl (setine/threonine kinase) n stanu:e, Noonan-like 0 Cl facies, hearing loss52 z

Irritability in ::j Costello syndrome Coarse facial features, HRAS (85-90%)2. n-74 HRAS (small G-protein) tvWd to moderate CNS abnormalities"

~ (rare)• deep palmar/planL1t BRAF (-4-6%)" BRAF (setine/ threonine kinase) mental 1\IR •, delay in young children, tli creases, papillomata, KRAS(7%)62 KRAS (small G-protein) language and motor seizures76 S' short stature, cardiac J.fEK1 (-2-3%)" .1\IEKl (tyrosine/serine/ development~ (f>

~ problems >2 threonine kinase) macrocephaly75• 76 en tij Cardia Facial Cutaneous "Noonan-like, '\vith BRAF(43-78%)'·s.77 BRAF (setine/threonine kinase) lvloderate to severe Obsessive behavior, sleep Aggression 79 w

syndrome (CFC, rare) bitemporal constriction, J1ffiK/ (1-11%)'· 11 1v!EK1 (tyrosine/setine/ .1\ffi, hypotonia, disturbance, failure to thrive, sparse hair, ulerythacma threonine kinase) marked delay in macrocephaly, CNS ophryogenes, cardiac MEK2 (6-7%)'· 77 MEK2 (tyrosine/setine/ language and motor abnormalities, seizurcs7B, 79

problems" threonine kinase) development78 KRAS {5-8%)'·" KRAS (small G-protein)

:c·

~ ?' lv

~ -...]

Table 1 (continued)

Disease (prevalence)

Prominent phenotypical characteristics

Hamartoma syndromesc

Tuberous Sclerosis Complex (TSC, 1:6000)

Bannayan-Riley­Ruvalcaba (BRR, rare)

Cowden Syndrome <rare>

Hamartomas, hypo­melanotic macules, facial angiofibromas, renal angiomyolipomas

Macrocephaly, hamartomas Qipotnas, hemangiomas), penile maculcsR2

Macrocephaly, 1\-lucocutaneous lesions, high frequency of various types of malignancies82

Genes e/o of cases associated with gene)b

TSCI (19%)"'

TSC2 (66%)'" r

PTEN (60%)83

PTEN (80-90%)"

Protein (function)

Hamartin (binding partner of Tubel1n)

Tuberin (RHEB-GAP)

PTEN (tyrosine phopshatase)

PTEN (tyrosine phopsharnse)

CNS featuresa

Very frequent (75-100%)

CNS abnortnalities, seizuresBI

Macrocephaly, developmental delay"'

Frequent (25-74%)

Bimodal IQ distribution; 50% normal IQ, 30% severe l\IR, specific deficits in attentional­cxccutive skills, memory and language, psychiatric disturbances including autistnl2

CNS abnormalities"

Less frequent (up to25%)

Subependyrnal giant cell astrocytomas•

SeizuresB-f

Learning disabilities, autisrn82

a Rare; up to a few hundred cases reported in literature; GAP: GTP-ase Activating Protein; GEF; Guanine nucleotide exchange factor; CNS; Central Nervous System; l'viR; mental

retardation (mild: IQ 50-69; moderate; IQ 35-49; severe; IQ <:: 34); ADHD; Attention Deficit Hyperactivity Disorder; l'viRI: Magnetic Resonance Imaging

b The - sign indicates that the mutation is reported in a subgroup of patients that is negati1'e for a combination of other mutations associated with the disease, without specifying the size

of the original population. In order to obtain an estimate of the prevalence of this mutation, we have corrected the percentage reported for the percentage in which these other mutations

are postulated to occur, as reported in this table.

' LEOPARD is "n acronym for the manifestations of this syndrome; multiple lentigines, electrocardiographic conduction abnormalities, ocular hypertelorism, pulmonic stenosis,

abnormal genitalia, retardation of growth, and sensorineural deafness.

dIn d1e majority of studies, Costello patients with mutations in genes other than HRAS are re-diagnosed to CFC syndrome.

' The Hamartoma syndromes compromise Tuberous Sclerosis Complex, Peutz-Jeghers syndrome, and the subgroup of the PTEN-hamartoma tumor syndromes, consisting of

Bannayan-Riley-Ruvalcaba, Cowden syndrome, Proteus syndrome and Lhermitte-Duclos disease. These latter group of diseases are all caused by germ line mutations in the PTEN

gene.83 However, because information on the cognitive phenotypes ofPeutz-Jeghers syndrome, Proteus syndrome and Lhermitte-Duclos disease is very limited, these syndromes are not

included.

'Calculated only for patients with a clinical diagnosis ofTSC.

Noonan, LEOPARD, and NF1 are associated with mild cognitive deficits. However, despite

low frequencies of mental retardation (IQ < 70) in NF1 and Noonan syndrome,12-15 ~40% of

the children require special education.12,13 In addition, even children with NF1 with a normal IQ

can still display specific deficits in multiple cognitive domains, including visual-spatial skills,

attention, executive functioning, and memory which puts them at risk for specific problems at

school or work_13, 14

In contrast to these relatively mild phenotypes, patients with Costello or CFC syndrome present

with high frequencies of mental retardation (Table 1). It is tempting to speculate that the

generally milder cognitive phenotypes in Noonan, LEOPARD, NF1 and NF1-like syndrome are

because of the fact that the causative mutations affect regulators of the RAS-ERI< pathway.

Mutations affecting the RAS, RAP and MEK (mitogen-activated and extracellular-signal

regulated kinase kinase) proteins, found in Costello syndrome and CFC, might have a stronger

effect on the output of the pathway (Figure 1). However, there are no data direcdy comparing

activity levels of RAS-ERI< signaling in brain tissue in the different disorders. Moreover,

mutations in the same gene can yield variable phenotypes, such that patients with identical

amino acid changes have been diagnosed with different syndromes (see Box 1 for striking

examples). Thus, the relationship between genotype and cognitive phenotype is still poorly

understood.

RAS-ERK signaling can modulate synaptic plasticity by regulating processes at both sides of the

synapse: at the presynaptic side it modulates neurotransmitter release (in the axon terminal of

the presynaptic neuron) and at the postsynaptic side, it controls protein synthesis (at the

dendritic spines of the postsynaptic neuron) (Figure 1, 2).

A presynaptic RAS-ERK pathway modulates neurotransmitter release

By changing the amount of neurotransmitter released from its axon terminal, the presynaptic

neuron can affect the strength of a synaptic connection (Figure 2). Several lines of evidence

suggest that the RAS-ERK pathway is involved in this process, which is probably activated by

binding of the neurotrophin BDNF (brain-derived neurotrophic factor) to the presynaptic

TRI<B (tyrosine receptor kinase type B) receptor. Bdnf mutant mice show a decrease in

neurotransmitter release, 16 whereas stimulation of the RAS-ERI< pathway by the application of

BDNF, as well as by expression of the active H-Ras(G12VJ (Harvey rat sarcoma viral oncogene

homolog) gene, results in an ERI<-dependent enhancement of neurotransmitter release. This is

achieved by ERI< phosphorylation of synapsin-I, a protein that binds to synaptic vesicles

3 8 ONCOGENES ON l\W MIND: ERKAND MTOR SIGNALING IN COGNITIVE DISEASES

containing neurotransmitters16. 17 (Figure 1,2). The presynaptic RAS-ERI-C signaling pathway is

not only controlled by neurotrophins. A recent study showed that also stress can induce

presynaptic changes via the RAS-ERK pathway. Activation of this pathway is induced by

corticosterone binding to the mineralocorticoid receptor, IS but it is still unclear how activation

of this receptor couples to RAS-ERK signaling.

BOX 1: WHAT'S IN A NAME? DIAGNOSING THE NEURO-CARDIO-FACIAL-CUTANEOUS SYNDROMES

A clinical or genetic diagnosis of a syndrome is invaluable to the affected patient and its parents in two

aspects. First, they can identify themselves with families with the same disorder, and second, they

expect to get a clear prognosis. However, the large overlap in phenotypes of the NCFC syndromes, the

desire to diagnose patients at a young age, even though the phenotype might still be obscure, and the

frequent lack of definite diagnostic criteria make it difficult to establish a diagnosis in an affected

patient. This is especially true for patients with overlapping characteristics of Noonan, Costello and

CFC syndromes.1° Now that many genes have recently been identified, would a diagnosis based on the

identified genetic mutation ensure a more accurate prognosis for the patient? Unfortunately, this is not

the case, because even patients with identical mutations often have highly variable phenotypes. For

example, identical mutations at D153V in KRAS were found in children diagnosed with Noonan

syndrome,'2 severe Noonan with CFC features,10 and CFC.3 Likewise, mutations at E501K in BRAF

are reported in patients with Noonan9 and CFC,3 and BRAF A2%P mutations in CFCJ and in Costello

(the latter rediagnoscd as CFC).63 This indicates that modifier genes, of which none are identified at

present, have an important role in shaping phenotypes in these syndromes.

Model organisms like mutant flies and mice are now used to identify these modifier genes.

Therefore, future research might lead to a novel classification system based on a 'fingerprint' of a large

number of selected genes that segregates patients on the basis of a certain prognosis (eg, malignancy

risk or cognitive function) rather than on a mutated gene or a syndrome diagnosis.

Expression of the active H-Ras(G12V) gene in a subset of neurons that form stimulating

synapses on their target neurons (excitatory neurons), resulted in enhanced synaptic plasticity

and improved learning in an ERl<:-Synapsin-I-dependent manner.17 This observation was

surprising, because even though most of the NCFC disorders are also characterized by increased

RAS signaling, the patients have learning deficits. The most probable explanation for this

apparent paradox is that increased RAS-ERl<:-Synapsin-I signaling in these diseases is mostly

restricted to inhibitory neurons. In contrast to excitatory neurons, inhibitory neurons form

repressing contacts on their target neurons. Indeed, Njt mice (Njt heterozygous knock-out

mice) show increased inhibitory transmission, which is probably mediated by enhanced release

of the main inhibitory neurotransmitter in the central nervous system (GABA; Gamma-

CH\PTER2 3 9

~

0

~ n 0 C) t:i z t:i en 0 z "' ~ i!:

~ i:Ii

~ ~ ~ >-1 0 ;r: "' c; z ;>­r z C)

z n 0

~ 3 51 s "' :;: <n t:i Cfo

Table 2: Mouse and rat mutants of genes associated with the RAS-ERK• and PI3K-MTOR signaling pathways and their phenotypes with respect to hippocampal functionh

Gene

RAS-GRFI

RAS-GRF2

.lj11GAP

HRAS

HRAS

KRAS

NRAS

NFI

BRAF

1'11EK1

1'11EK1

Mutation Phenotype

Hippocampal-dependent Learning<

Ras-Gifl homozygous knock- Impairments in some spatial out mouse learning patadigms,86 intact

performance in others22

Ras-Grf2 inducible homozygous Not known knock-out mouse23

SJ''l'flP heterozygous knock-out Impaired spatial learning"" mouse

H-Ras homozygous knock-out Not known mouse

Forebrain and excitatory neuron- Enhanced spatial learning specific constitutively active H-Ras(H-RasGI2V) mouse mutant17

K-Ras heterozygous knock-out Impaired spatial learning mouset'J

N-Ras heterozygous knock-out Intact spatiallearrting mouse I?

NJI heterozygous knock-out Impaired spatinllearning" mouse (see Box 2)

Forebrain and excitatory neuron- Impaired spatial learning specific B-Ra/homozygous knock outmouse'J3

Neuron-specific dominant- Impaired spatial learning negative Mekl mutant"'

Forebrain and exdtatory neuron- Impaired long term spacial specific domin.'\tlt-negative iHekl memory mouse murant:95

Synaptic plasticity"

Impaired L1D,23 intact LTP in some p.rotocols22 and slight iropairment in othecsn

Impaired LTP and decreased presynaptic plasticity

Impaired L 'fP29, ss

Enhanced L TP in some protocols,90 intact LTP in others;88 increased N1\'IDA­.rcceptor mediated responses90

Enhanced L TP and increased presynaptic plasticity

Impaired L TP

Not known

Impaired LTP and increased GABA-mediated inhibition'"

Impaired LTP

Not known

Impaired late phase LTP

Molecular signaling

Intact NMDA-receptor induced ERK phosphorylationn

Decreased NMDA-receptor induced ERK phosphorylation

Increased basal ERK and 1'-IEK phosphorylation, increased Nl'viDA­receptor induced ERI( phosphorylationBB

Increased phosphorylation of NR2A and NR2B subunits of the NMDA receptor;"' intact basal ERK and MEK phosphorylationss

Increased basal ERI< and SYNI phosphorylation, intact basal AKT phosphorylation

Not known

Not known

Increased basal ERI< and CREB phosphoryL•tion;so. "' intact basal AKT phosphorylation"'

Intact basal ERK phosphorylation; decreased ERI< phosphorylation after a learning paradigm

Not known

Decreased protein synthesis upon LTP inducing stimuli; decreased ERI(, S6 and eiF4E phosphorylation upon LTP inducing stimuli and after a learning paradigm

Morphology

No apparent changes in brain morphology22

No apparent changes in brain motphology87

Increased number of AlvJP A­receptor clusters in neuronal cultures of homozygous knock­out mlce89

No apparent changes in brain morphology"· 90

Increased amount of ncurotmnsmitters ready for release (docked vesicles)

Not known

Not known

~Mild astrogliosis'"'

Not known

No apparent changes in brain morphology

Not known

'"'· '' :-j ~

'"

..j;:. ,....

Table 2 (continued)

Gene

ERK1

ERK2

PIJK

PTEN

TSC1

TSC2

TSC2

Mutation

Erkl homozygous knock-out mouse

--Knock-down mouse mutant with a 20-40% reduction in Erk2 exprcssion9s

p58a (regulatory subunit of PiJk) knock-out mouselS

Nfousc mutant wid1 homozygous Ptm deletion in limited neuronal populations (including hippocampus)

Tir1 heterozygous knock-out mouseJs

TH2 hctcrozrgous knock-out rat

Tsr2 heterozygous knock-out mouse~9

Phenotype

Hippocampal-dependent Learning<

Enhanced spatial learning in some protocols,96 .intact performance in others97

Impaired spatial learning

Impaired spatial learning

Impaired spatial lcarning36

Impaired spatial learning

Intact spatial lcarning99

Impaired spatial learning, was rescued by treatment with Rapamycin

Synaptic plasticity<

Impaired LTP in some protocols,96 intactLTP in others%· 97

Nor known

Not known

Impaired basal transmission andLTP''

Not known

Impaired LTP and LID and increased presynaptic plasticity••

Lower threshold for late phase L TP, was rescued by treatment with Rapamycin

Molecular signaling

Both increased" and intacl"' ERK2 signaling reported.

Not known

Not known

Increased basal AKT, 1\ITOR and S6K phosphorylation"

Not known

Intact basal ERK phosphorylation, increased ERIC phosphorylation upon L TP inducing stimuli

Increased basal S6 phosphorylation, was rescued by treatment with Rapamycin

Morphology

No apparent changes in brain morphology"· 97

No apparent changes in brain morphology

Decreased synaptic densit}'

Hypertrophy of cell soma, ectopic dendrites and axonal tracts and increased spine densitylli. 37

No neuronal abnormalities, no lesions by MRI

Adult animals arc free of cerebral hamartomas, aged animals develop them at a slow rate27

No apparent changes in brain morphology

'RAS, rat sarcoma viral oncogene homolog; ERIC, extracellular signal regulated kinase; PI3K, phosphatidylinositol3-kinase; MTOR, mammalian target of rapamycin; LID,

long-term depression; LTP, long-term potentiation; Nl'viDA, N-methyl-n-aspartate; MEK, mitogen-activated and extracellular-signal regulated kinase kinase; Al'viP A, alpha-amino-

3-hydroxy-5-methyl-4-isoxazolepropionic acid; GABA, gamma-aminobutyric acid; CREB, cAl'viP response element-binding.

b Only rodent mutants in the direct RAS-ERK and PI3K-MTOR routes in which hippocampal function is specifically tested are presented.

' See Glossary.

aminobutyric acid). This increased inhibitory transmission seems to direcdy cause the

impairments in plasticity and learning in these mutants19 (Box 2). This is an interesting example

of how a similar modification of the RAS-ERK pathway, can generate opposite systems-level

outcomes by affecting two different types of neurons. However, it is not yet clear how the NF1

mutation affects RAS-ERK signaling preferentially in inhibitory neurons.

BOX 2: TREATING COGNITIVE DEFECTS IN NFl- LOST IN TRANSLATION?

Similar to NFl patients, Njl heterozygous knockout mice have problems in learning and attention.19· 50

In addition, they show deficits in synaptic phsticity.19 Importantly, these deficits can be rescued by

genetically reducing the level of N-RAS or K-RAS, suggesting that learning and plasticity deficits in

Njl mice are caused by enhanced RAS signaling.1'

RAS activity is critically dependent on its association to membranes, for which it requires

the post-translational addition of a farnesyl or geranylgeranyl anchor. Statins decrease the synthesis of

cholesterol, farnesyl and geranylgeranyl by inhibiting HMG-CoA (3-hydrm:y-3-methylglutaryl­

coenzyme A) reductase, the rate-limiting enzyme in the mevalonate synthesis pathway. Interestingly,

treatment of Njl mice or flies with statins cures their learning deficits. so. 54 Statins are prescribed widely

to treat hypercholesterolaemia, and have an excellent safety profile in adults and children. Therefore,

the effect of simvastatin on cognitive functioning has recently been investigated in a randomized,

placebo-controlled trial, involving 62 children with NF1.6* Outcome measures included

neuropsychological tests, MRI analysis and a neurophysiological test (measuring eye-hand movement

control). Unfortunately, a three-month treatment resulted in a significant improvement in only one out

of nine neuropsychological outcome measures, when compared to the placebo group. Several factors

could have atrtibuted to these disappointing results. First, it is conceivable that reversing deficits in

higher cognitive functions in humans is far more difficult than reversing cognitive deficits in mice.

This could be due to the greater complexity of the human brain. Second, the statin concentration that

was reached in the human brain, could have been significant lower than in mice. This could be due to

differences in metabolism, or to differences in blood-brain barrier permeability. Third, there was a

large placebo or re-test effect, which brought 3 of the 9 neuropsychological outcome measures back to

normal values. Since statins did not improve cognitive function in wild-type mice, it is possible that a

ceiling effect was reached for these measures. Finally, it can not be excluded that the tests were not

sensitive enough to capture a real improvement (see also Box 3). Because of all these biological and

methodological issues, trials involving a longer treatment are now initiated. This would allow the brain

more time to undergo changes, and would diminish the placebo and re-test effect by increasing the

time in between testing moments. Moreover, a longer treatment would allow inclusion of real-life

measures such as school performance.

The postsynaptic RAS-ERK pathway is an important signal integrator

Synaptic strength is not only controlled by regulating the amount of neurotransmitter release. In

fact, most of the changes taking place during memory formation occur on the postsynaptic side

4 2 ONCOGENES ON !v!Y MIND: ERKAND MTOR SIGNALING IN COGNITIVE DISEASES

of the synapse. Influx of calcium ions through NMDA (N-methyl-D-aspartic acid) receptors is

the pivotal trigger to initiate the process of synaptic strengthening, which can be measured in

vitro (then referred to as LTP; long-term potentiation), and which is an absolute requirement for

learning and memory. The RAS-GEFs (guanine nucleotide exchange factors) are recognized as

major connectors between calcium ions and RAS-ERK activation, as they associate with

NMDA receptors, and are activated by the influx of calcium ions through these receptors.zo. 21

Genetic studies suggest that RAS-GRF2 (guanine nucleotide-releasing factor) is the main GEF

that drives RAS-ERJ<.-dependent synaptic strengthening.22. 23 However, an increase in calcium

can also activate ERJ<. through CaMKII (calcium/ calmodulin-dependent protein kinase 2)

mediated inactivation of SynGAP (synaptic RAS GTPase activating protein), a negative

regulator of RAS signaling24 (Figure 1). Besides calcium influx, the postsynaptic RAS-ERK

pathway can also be activated by BDNF binding to the TRJ<.B receptor, by the activation of()­

adrenergic receptors and by a, more indirect, cAMP-PKA (protein kinase A)-dependent

pathway2S-27 (Figure 1). Hence, the postsynaptic RAS-ERJ<. pathway serves as a major signal

integrator to control synaptic plasticity.

The postsynaptic RAS-ERK signaling pathway has many targets

How does the postsynaptic RAS-ERK pathway control postsynaptic plasticity? By changing the

number of AMPA (alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) receptors in the

cell membrane, the postsynaptic neuron directly controls its sensitivity to glutamate and its

probability to flre an action potential (Figure 2). Both over-expression of an active form of RAS

in hippocampal neurons and silencing of SynGAP lead to an ERJ<.-dependent increase in the

amount of AMP A receptors in the postsynaptic membrane.2B, 29 This suggests a direct link

between postsynaptic ERK signaling and AMP A receptor dynamics.

Protein synthesis is an absolute requirement to convert transient changes in synaptic strength

into stable, long-lasting connections, and hence in stable memories. When measured in vitro, this

phase of synaptic strengthening is referred to as late-phase LTP (L-LTP). ERJ<. signaling plays a

crucial role in controlling protein synthesis by regulating both transcription and translation

events. One of the targets of the RAS-ERJ<. pathway is the transcription factor CREB (cAMP

response element-binding), which is important for memory formation.26 The regulation of

CREB seems to be sensitive to the application of BDNF30 and its activation is dependent on

several kinases downstream of ERJ-(26 (Figure 1). One of these, RSK2 (ribosomal S6 kinase 2),26

is associated with the X-linked Coffm-Lowry syndrome (OMIM 303600) and patients with this

disease present with mental retardation. Notably, the CBP (CREB-binding protein) gene, which

CR\PTER2 4 3

encodes an essential transcriptional co-activator of CREB, is mutated in Rubinstein-Taybi

Syndrome, a disease characterized by severe mental retardation (OMIM 180849), again stressing

the importance of proper ERK-dependent signaling in cognitive function.

Besides its roles in transcription, ERI< signaling also controls translation in concert with the

MTOR signaling pathway and more directly by activating the MNK (mitogen activated protein­

interacting kinase) isoforms, which in turn activate eiF4E (eukaryotic initiation factor 4E)31

(Figure 1). Collectively these studies indicate that RAS-ERI< signaling plays a critical role in

several major aspects of synaptic plasticity.

Cognitive deficits arising from genetic impairments in PI3K-MTOR signaling

Tuberous Sclerosis Complex (TSC) and the PTEN (phosphatase and tensin homolog)­

hamartoma tumor syndromes, a group of clinical entities all resulting from germ-line mutations

in PTEN, are classified as the Hamartoma syndromes (Table 1). The PTEN-hamartoma tumor

syndromes present with mental impairments, however because of their rare nature detailed

descriptions on the cognitive phenotypes are sparse. The cognitive profile of TSC patients is

remarkably variable, with half of the patients having a normal IQ and about 30% an IQ below

20.32 Similar to NF1, specific deficits in attention, executive functioning, memory and language

are also common in TSC patients with a normal IQ.32 The variation in cognitive abilities can in

part be explained by differential effects of TSC1 (Tuberous Sclerosis Complex 1 gene) versus

TSC2 (Tuberous Sclerosis Complex 2 gene) mutations (TSC2 mutations tend to aggregate with

more severe cases of mental retardation), and the abundance and localization of brain

hamartomas and the presence and severity of epilepsy (!'able 1).32 The mutational spectrum of

the two TSC genes is very broad, complicating research into a possible contribution of modifier

genes to the variability in phenotype. Hence, no modifier genes that affect cognitive function

have been identified. However, polymorphisms in the Interferon-y gene33 and in the gene

encoding the DNA repair agent 8-oxoguanine glycosylase 1 (OGG1)34 modulate susceptibility to

renal angiomyolipomas in TSC patients, pointing to a role for modifier genes in the

phenotypical variability of TSC.

PI3K-MTOR signaling controls protein translation

Rodent models have been developed for both TSC as for the PTEN-hamartoma tumor

syndromes, and studies in these mutants reveal an essential role of PI3K-MTOR signaling in

learning and memory35-38 (Table 2). Both Pten homozygous knock-out mice and Tsc1 and Tsc2

heterozygous knock-out mice have impaired learning and show deficits in synaptic plasticity

4 4 ONCOGENES ON MY MIND: ERK AND MTOR SIGNAI.ING IN COGNITIVE DISEASES

(Table 2).36-39 However, unlike the patients with PTEN-hamartoma tumor syndromes, Pten

mouse mutants have severe disruptions in brain architecture,37 which is probably related to these

mice carrying a homozygous rather than a heterozygous deletion. It is not clear whether the

learning deficits are secondary to these developmental brain abnormalities or the direct result of

aberrant neuronal plasticity in the absence of PTEN. By contrast, a heterozygous Tsc1 mouse

mutant showed learning impairments in the absence of brain pathology or seizures, implying

that the TSC proteins have a direct role in synaptic plasticity.38

Like the RAS-ERI< pathway discussed above, the MTOR pathway is involved in the protein

synthesis-dependent phase of synaptic strengthening. Rapamycin, a selective inhibitor of

MTOR, specifically impairs this phase of synaptic strengthening and causes long-term memory

deficits.4D. 41 Interestingly, all components of the PI3K-MTOR pathway and the complete

translation machinery are present in dendrites,41 suggesting that this pathway controls protein

translation near the activated synapse (designated as 'local' protein translation). Indeed, BDNF

stimulation is found to initiate MTOR-dependent protein translation in isolated dendrites.42

Besides BDNF, activation of NMDA- and ~-adrenergic receptors can also induce MTOR

signaling.zs, 43 MTOR drives local protein translation through phosphorylation of its downstream

targets, which include 4E-BP1 (eukaryotic translation initiation factor 4E-binding protein) and

S6K (S6 kinase)44 (Figure 1). In addition to its important role as an initiator of local protein

synthesis, MTOR can also suppress local translation of certain proteins, among which is the

Kvl.l potassium channel.4S

As expected based on their increased MTOR signaling, Tsc2 mutant mice show increased

phosphorylation of S6 ribosomal protein, which is involved in protein translation (Figure 1).

Consequendy, a relatively weak stimulus is sufficient to recruit the protein synthesis-dependent

phase of synaptic strengthening in these mutants. Paradoxically, this causes a learning deficit

rather than a learning enhancement, probably because of inappropriate storage of unrelated or

unprocessed information_39

It remains to be elucidated which dendritically targeted mRNAs are specifically regulated by

MTOR, and how this couples to synaptic strengthening. However, a direct link has been

established between PI3K signaling and AMPA receptor insertion, suggesting that this might be

one of the main mechanisms by which MTOR signaling drives long lasting synaptic changes.46

Taken together, these studies imply that a crucial balance of MTOR signaling is required to

control neuronal protein translation, which is essential to long-term synaptic changes.

CH.\J"J'ER 2 4 5

ERK and MTOR signaling in autism

Thus far, we have focused on the roles of ERI<. and MTOR signaling in cognitive function;

however, behavioral problems are also commonly associated with both the NCFC and the

Hamartoma syndromes. Evidence for a relationship with autism is somewhat limited for both

NF147. 48 and Noonan syndrome48, but autism is certainly a prominent characteristic of the

Hamartoma syndromes (fable 1). One half of the TSC patients have autistic features 32.

Conversely, mutations in the TSC genes are found in 1% of autistic individuals and PTEN germ

line mutations are found in as many as 17% of patients presenting with both autism and

macrocephaly.48 These are strikingly high percentages in light of the still obscure genetic

knowledge on autism.

Mouse models for the PTEN-hamartoma tumor syndromes and TSC are found to recapitulate

the social withdrawal phenotype as seen in autistic individuals.36• 38 Taken together, the high

incidence of autism in Hamartoma syndromes patients and the autistic phenotypes in mouse

models for these syndromes make clear that aberrations in the PI3K-MTOR pathway can cause

molecular and cellular changes that lead to autistic behavior. Thus, this pathway might be a

major player in causing autism. However hampered by a lack of knowledge of the brain areas

involved in autism, insight is limited into the exact mechanisms leading to autism upon

enhanced PI3K-MTOR signaling.

Cognitive impairments are related to reversible changes in signaling rather than gross

brain abnormalities

Structural brain abnormalities and seizures are part of the phenotypic spectrum of the NCFC

and Hamartoma syndromes (fable 1). It could be argued that the cognitive deficits develop only

secondary to these brain abnormalities. However, evidence for this idea is limited and contested.

First, although infantile spasms are associated with a poor cognitive outcome in TSC,32 clinical

studies fail to show consistent data on a correlation between MRI abnormalities and cognition

in TSC and NF1.49 Second, cognitive impairments in most of the mouse models for NCFC and

Hamartoma syndromes occur in the absence of structural brain abnormalities as seen in patients

(Table 2). Third, as outlined in previous sections, the cognitive impairments found in these

mouse models seem to arise from disturbances in the balance of neuronal signaling, because

treatments with drugs specifically targeting these signaling disturbances can rescue both the

cognitive deficits as the impairments in synaptic plasticity in mouse models for TSC and NF1

(Box 2).39. so Interestingly, recent results show that, even though epilepsy correlates with poor

cognitive outcome in TSC, this symptom can also be directly attributed to disturbed MTOR

4 6 ONCOGENES ON MY MIND: ERK Al'ID MTOR SIGNAI.ING IN COGNITIVE DISEASES

signaling, and can be rescued with Rapamycin.51 These findings suggest that the cognitive

impairments are not caused by irreversible developmental abnormalities of the brain, but can be

attributed to reversible changes in signaling.

Treating cognitive genetic disease -lessons from cancer

Cancer research has generated a wealth of knowledge on how to interfere with RAS-ER!<:. and

PI3K-MTOR signaling. By exploiting this knowledge we might be able to reverse the cognitive

deficits associated with the NCFC and Hamartoma syndromes. However, there are several

important aspects that have to be considered when using oncology drugs to treat cognitive

deficits. First, animal studies suggest that both increased and decreased ERl<:. or MTOR

signaling result in cognitive impairments, indicating that a strict balance is required (fable 2).

This is in marked contrast to tumors associated with these diseases, which are always caused by

up-regulation of the ERl<:. or MTOR pathway, and often require an additional second hit

affecting the other allele (loss of heterozygosity) to become oncogenic. 52 Thus, treating

cognitive deficits requires considerably more careful dosing than the treatment of cancer.

Importantly, this implies that high doses of these drugs, as used in cancer treatment, might

negatively affect cognitive function, which is of considerable concern (Box 3). Second, side­

effects are acceptable in treating a life-threatening tumor, especially if the treatment is short.

However, the treatment of cognitive deficits would probably be life-long and therefore requires

an exceptionally good safety profile. Finally, many of the small molecule inhibitors used in

cancer treatments are specifically designed to not be able to cross the blood-brain barrier, which

makes them unsuitable to treat cognitive disorders.

Treatment ofNCFC .ryndromes with inhibitors of the RAS-ERK pathwqy

Inhibiting RAS activity is a potential treatment mechanism for the cognitive impairments in the

NCFC disorders (I•'igure 1). RAS activity can be diminished by attacking its Achilles' heel: its

requirement to be post-translationally modified (Box 2). Both farnesyl transferase (FTase)

inhibitors and statins can reduce RAS signaling in this manner, but although they show anti­

proliferation effects in vitro, their success in treating cancer as a monotherapy has been limited

(reviewed in 53). Nevertheless, it is probable that the amount of RAS inhibition required to treat

cognitive deficits is significantly lower than for tumor regression. Indeed, both FTase inhibitors

and statins were sufficient to rescue cognitive and plasticity deficits of Nf1 mice,19, so and more

recently to rescue learning impairments in Nf1 mutant flies,54 suggesting an evolutionary

conserved mechanism. Despite these flndings, a clinical trial assessing the effects of simvastatin

in NF1 patients showed little effect (Box 2).

CH \PThR2 4 7

It has to be noted that the in vitro antiproliferative effects of both FTase inhibitors and statins

cannot solely be ascribed to their ability to interfere with RAS signaling.ss Therefore, we cannot

rule out that the rescue of the learning deficits of Nf1 mice is the result of other mechanisms.

For instance, statins might reduce the synthesis of neurosteroids, because the rate-limiting step

of steroid synthesis is the conversion of cholesterol to pregnenolone. Because neurosteroids

directly activate the inhibitory GABA-A receptor, 56 reduction of neurosteroid levels might help

to decrease the enhanced inhibition mediated by GABA-A signaling, as observed in Nf1 mice.19

BOX3: COGNITIVE FUNCTION AND CHEMOTHERAPY: THE CHEMOBRAIN

The increasing number of patients surviving cancer has aroused interest in how chemotherapy affects

quality of life. Patients receiving conventional chemotherapy that causes DNA damage and cell death

(e.g. platinum compounds) often report transient or even persistent cognitive impairments across

various domains including working memory, executive function and processing speed (reviewed in ref.

6'). However, the precise impact of chemotherapy on brain function is a matter of debate for two

major reasons. First, some of the studies reported a discrepancy between self-reported problems and

objective neuropsychological tests, with no clear correlation between these two measures. 66• 67 Second,

most studies are cross-sectional studies; hence, cognitive performance of the subjects before treatment

is not known. Recently, several prospective (longitudinal) studies on this topic have been published,

and although most studies suggest that chemotherapy has an impact on cognitive function, it does not

seem as dramatic as reported by some earlier cross-sectional studies (for a review, see Ref 66). Possibly,

the effects in these prospective study designs were smaller, because patients are repeatedly assessed

with similar tests, which can result ill practice effects that might mask a real . cognitive decline.

However, one illteresting aspect that was noted in several of these prospective studies was a greater

than expected incidence of cognitive problems in these patients even before initiation of chemotherapy

(see Ref 65 and references therein). Although several factors including psychological factors (e,g. stress,

anxiety, depression after being diagnosed with a life-threatening disease) and biological factors (e.g.

cytokine elevation) could cause this pretreatment deficit in cognitive functioning, it is tempting to

speculate that certain polymorphisms in the genes the function in the RAS-ERK or PI3K-MTOR

pathways result in ill creased cancer susceptibility as well as decreased cognitive function.

Because of the DNA-damaging nature of conventional chemotherapies, neuronal cell

death is more likely to be an important mechanism underlying the induced cognitive problems than

direct interference with synaptic plasticity. By contrast, the novel chemotherapies that are based on

small molecule inhibitors directed against proteins ill the pathways discussed in this review can directly

impede synaptic plasticity. Thus, provided that they can cross the blood-brain barrier, they might

severely affect cognitive function. For instance, MEK and MTOR illhibitors are found to affect

cognitive function in wild type mice.4°. 68 Therefore, substantial animal and clinical studies will be

required to assess the short-term and long-term effects of these new cancer treatments on cognitive

function.

4 8 ONCOGENES ON !V1Y MIND: ERKAND MTOR SIGNALING IN COGN!TIVE DISEASES

The RAS-ERI< pathway can also be targeted with several newly developed small molecule

inhibitors of B-RAF (v-raf-1 murine leukemia viral oncogene homolog Bl) and MEK. Cancer

trials with these inhibitors are underway (reviewed in sry. However, this first generation of small

molecule inhibitors is probably not suitable for treating cognitive deficits, because oflow blood­

brain permeability and significant side effects.

Treatment of Hamartoma !Jndromes with inhibitors of the MTO R pathwqy

The MTOR inhibitor Rapamycin (Sitolimus), applied as immunosuppressant in organ transplant

patients, has already been successfully used to treat astrocytomas and angiomyolipomas in TSC

patients,33, ss, 59 Rapamycin is also shown to have anti-proliferative effects in patients with brain

tumors caused by reduced PTEN activity.60 A recent study revealed that rapamycin can reverse

the cognitive deficits and aberrations in synaptic plasticity in Tsc2 mutant mice.39 In addition, a

clinical trial to measure the effect of rapamycin treatment on renal hamartomas in TSC patients

in conjunction with cognitive function (memory and executive skills) as secondary outcome

measure is underway (NCT00490789; http://clinicaltrials.gov). Ideally, all future Rapamycin

trials in TSC patients should include some measures to assess cognitive itnprovements and

quality of life. The success of the cognitive itnprovements (if any) should then be carefully

weighed against the drawbacks associated with a long-term treatment with rapamycin.

Finally, like RAS, the TSC target protein RHEB (RAS homolog enriched in brain) is crucially

dependent upon farnesylation. This suggests that FTase inhibitors and statins might also help to

treat the cognitive deficits in the Hamartoma syndromes.61

Concluding remarks

Here, we have emphasized the itnportance of the oncogenic RAS-ERK and PI3K-MTOR

signaling pathways in cognitive functioning by focusing on the cognitive deficits associated with

the NCFC and Hamartoma syndromes. There is a strong connection between genetic alterations

in components of these pathways and cognitive dysfunction. Because of pioneering studies in

cancer research, these signaling routes are very well characterized and rapid progress has now

been made to understand theit role in neuronal function. Animal studies revealed that the

neuronal RAS-ERK and PI3K-MTOR pathways modulate neurotransmitter release, control

synthesis of proteins requited for stabilizing synaptic changes, and regulate receptor properties

and dynamics. These processes play a pivotal role in synaptic plasticity, requited for proper

cognitive function. Recent targeted treatments in animal models of NCFC and Hamartoma

CH,\PTER 2 4 9

syndromes, using drugs designed for cancer treatment have been successful, and will

undoubtedly stimulate the initiation of many clinical trials.

BOX 4 AREAS FOR FUTURE RESEARCH

Investigating to what extent the mechanisms underlying the more severe neuro-cardio-facial­

cutaneous (NCFq disorders are different from the mechanisms underlying the NCFC disorders

with only mild cognitive deficits.

Investigating how certain mutations only affect a subclass of neurons (eg. Neurofibromatosis

type 1 affects predominantly inhibitory neurons).

Identification of the mRNAs whose translation is controlled by mammalian target of rapamycin

(MTOR) and defining which are crucial for causing the cognitive deficits in the Hamartoma

syndromes.

Do the treatments, which can rescue cognitive functioning in mouse models of the Hamartoma

syndromes, also rescue their autistic phenotypes?

Compelling proof that the treatments that can rescue the cognitive deficits in mutant mice are

also effective and safe in patients.

Identification of small-molecule inhibitors of the e>:tracellnlar signal regulated kinase (ERK) and

MTOR pathways with minimal side-effects and that can cross the blood-brain barrier efficiently

so that they can be used to treat cognitive deficits.

GLOSSARY

AMPA RECEPTOR: Alpha-arnino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; Subtype of

the glutamate receptor, mediates fast excitatory synaptic transmission in the central nervous system.

ANGIOMYOLIPOMA: Benign kidney tumor composed of an abnormal collection of blood vessels

(angio), smooth muscle (myo), and fat (lipoma). Found in 70-80% ofTSC patients.

AsTROCYTOMA: Benign brain tumor composed of undifferentiated, dysfunctional glial cells. Found in

10-20% ofTSC patients.

AUTISM: A developmental disorder characterized by a triad of symptoms: a qualitative impairment in

social interaction, qualitative impairments in communication, and restricted, repetitive and stereotyped

patterns of behavior.

DENDRITE: A neuronal process arising from the cell body that receives synaptic input. From the

viewpoint of a specific synapse, this dendrite lies on the postsynaptic side.

ExCITATORY NEURON: A neuron that forms stimulatory contacts on its target neurons, and thereby

increases their pro babiliry to fire. Glutamate is the most common neurotransmitter released by

excitatory neurons.

5 0 ONCOGENES ON MY MIND: .ERK/\ND MTOR SIGNALING IN COGNITIVE DISEASES

GLOSSARY (CONTINUED)

GABA: Gamma-aminobutyric acid; Most abundant inhibitoty neurotransmitter in the central nervous

system.

GLUTAMATE: Most abundant excitatoty neurotransmitter in the central nervous system.

HAMARTOMA: A benign tumor-like growth consisting of a disorganized mixture of cells and tissues

normally found in the area of the body where the growth occurs.

HIPPOCAMPUS: Part of the brain essential for memoty formation. In rodents its function is typically

assessed with maze-tasks.

INHffiiTORY NEURONS: A neuron that forms inhibiting contacts on its target neurons, and thereby

reduces their probability to fire. In the central nervous system, GABA is the most common

neurotransmitter released by inhibitoty neurons.

LTD/LTP: Long-term depression/Long-term potentiation; An in vitro measure of synaptic weakening

and strengthening, respectively. LTP can be subdivided in an early phase, (1-2 hours after LTP

induction) requiring posttranslational changes only, and a late phase, which requires the synthesis of

new proteins. The protein synthesis-dependent phase of synaptic strengthening is required for long

term memoty formation.

MENTAL RETARDATION (MR): The combination of an IQ <70 (normal IQ is 100 ± 15) with

signi£cant limitations in at least two areas of adaptive behavior (e.g. communication, daily living skills

or social skills), apparent before the age of 18. An IQ of 69-50 is defined as mild; 35-49 as moderate;

20-34 as severe, and <20 as profound MR.

NEUROTROPHINS: Family of proteins, which are important for neuronal survival in the developing

brain, and play a role in synaptic plasticity in the mature brain.

NMDA RECEPTOR: A subtype of glutamate receptor. Mediates calcium inflm during L TP induction.

PLASTICITY (SYNAPTIC/NEURONAL): The ability of neurons to change the strength of synaptic

contacts or their excitability. These processes are required for memoty formation.

POSTSYNAPTIC: The side of the synapse on the dendrite where the receptors are located which are

receptive to the released neurotransmitter molecules.

PRESYNAPTIC: The side of synapse on the axon tertninal where neurottansmitter molecules are

released, which convey signals to the target cell.

ULiPTLR.?. 51

FIGURE 1: OVERVIEW OF NEURONAL RAT SARCOMA VIRAL ONCOGENE HOMOLOG (RAS)- EXTRACELLULAR SIGNAL

REGULATED KINASE (ERK) AND PHOSPHATIDYLINOSITOL 3-KJNASE (PI3K)- MAMMALIAN TARGET OF RAPAMYCIN

(MTOR) SIGNALING PATHWAYS AND THE ASSOCIATED NCFC AND HAMARTOMA SYNDROMES

The RAS-ERK and PI3K-MTOR pathways are both regulated by the activity of small GTP-binding proteins (G­

proteins), RAS and RHEB, respectively (accentuated in the figure with a thick border). These small G-proteins can reside

in two different states: a GTP (guanosine triphosphate)-bound active state and a GDP (guanosine diphosphate)-bound

inactive state. Their activity level is determined by interactions with GAP (GTPase activating protein) and GEF (Guanine

exchange factor) proteins. The different GEF proteins promote the exchange of GDP for GTP, leading to enhanced

activity, while GAP proteins catalyze the hydrolysis of GTP to GDP, leading to suppression ofactivity.

Activation of RAS is initiated by calcium influx through N-methyl-D-aspartate (NMDA) receptors, which in tum activates

RAS-guanine nucleotide-releasing factor (RAS-GRF2) and inactivates synaptic RAS GTPase activating protein

(SynGAP).Z0-24. 26.29 RAS can also be activated after mineralocorticoid receptor (MR) activation by corticosterone (cort),

~-adrenergic receptor (~-AR) activation by noradrenaline (NA) or by brain-derived neurotropic factor (BDNF) binding to

the tyrosine receptor kinase type B (TRKB) receptor, which initiates RAS signaling by activating the GEF protein: Son of

Sevenless, Drosophila, homologue 1 (SOS1), a GEF-homolog.16·18 Src homology protein 2 (SHP2) stimulates this activation

in as yet undefined ways. Active RAS activates RAP, which induces a phosphorylation cascade ultimately leading to

activation of ERK and its downstream targets. At the presynaptic side these include Synapsin-I (Syni), which modulates

neurotransmitter release. At the postsynaptic side, ERK activates ribosomal S6 kinase 2 (RSK2) and myocardial Snf1

(sucrose nonfermenting 1 )-like kinases 1, 2 (MSK 1, 2), which in turn activate the transcription factor cAMP response

element-binding (CREB). Also, ERK activates mitogen activated protein-interacting kinases 1, 2 (MNK1, 2), which signal

for translation. Also, ERK direcdy modulates the dynamics of ion channels including the potassium channel Kv4.2. All

these processes all influence the strength of the synapse, essential to learning and memory formation. MTOR is a major

controller of dendritic translation through its downstream targets S6 kinase (S6K) and eukaryotic translation initiation

factor 4E-binding protein 1 (4E-BP1). MTOR is driven by RHEB activity, which is increased by inhibition of Tuberin­

GAP (TSC2) activity. This could occur direcdy by RAS-PI3K signaling or indirecdy through RAS-ERIC signaling. Dashed

lines represent interactions of which not all details are elucidated at present.

The genes mutated in the different syndromes are depicted in the dark boxes, with the name of the

syndrome(s) next to them. The genes shown in the light boxes are not (yet) associated with a syndrome, but are plausible

candidate genes for cases in which no mutation is identified. The plus and minus signs indicate whether the identified

mutations up- or down-regulate ERK or MTOR signaling (based on in vitro assqys). Note that the vast majority of the

mutations encountered in the NCFC and Hamartoma syndromes lead to enhanced ERK or MTOR signaling.

Abbreviations: SPRED1: sprouty related EVH1 domain containing protein 1, PIP3: Phosphatidylinositol (3,4,5)­

trisphosphate, PIP2: phosphatidylinositol (4,5)-bisphosphate, PDK1: Pyruvate dehydrogenase kinase, isozyme 1, eEF2-K:

eukaryotic elongation factor 2 kinase AKT: v-akt murine thymoma viral oncogene homolog, S6K: S6 kinase, eiF4B/E:

eukaryotic translation initiation factor 4 B/E, eEF2-K: eukaryotic elongation factor-2 kinase.

5 2 ONCOGENES ON MY MIND: ERKAND MTOR SIGNALING IN COGNITIVE DISEASES

g ~·

~ ~

'" "Tl ..... CIQ g

(..J1 (b

~ .......

lsoprenyl ~~; anchor ~-"If/

~ ~ a.

Cell membrane

I 4!1 I

GO~: I ;;aL I I I 'T /f\1 8 . £@]I I" (

/,) I ~ J.#$ . " " .. ~-~:· . PIP2 •

it,;) _ ~ "0

"'" ~. ( ·· )_ •. · ··· EN-hamartoma

N'l .Jij.""+- . - ~ 1,~ [Fiffil '"m" ''"":;m~ C Noonan,CFC, r::;;~__.fA~ & c"'"'" ~ \:::/ oo

. -p ---J _l RHEB ) "" - "' 1 < ''"' Syndron~ •. .Noonan, CFC, • ( ~,'.! ~ ! Costello \ ' T

Noo"' c

"""',., "'~ ~'· •

Tuberous

A'esynaptic side

8

omplex

Noonan,CF~ C Costello

~ rERK\ -~ .....

Postsynaptic side Postsynaptic side

.® Coffin Lowry

~· Rubinst.~REBJ Taybi . m Postsynaptic side

®8 '..~

(e1F4E]

® !

Neurotransmitter ion channel modulation release Transcription Translation

FIGURE 2: THE OUTPUT OF ERK AND MTOR SIGNALING IN MITOTIC CELLS AND NEURONS

(a) In non-neuronal, mitotic cells, extracellular signals such as growth factors and cytokines induce proliferation,

differentiation and cell cycle progression via activation of ERK and MTOR pathways. However, neurons (b) are mostly

post-mitotic, and the ERK and MTOR pathways are recruited for a process called synaptic plasticity, important in

memory formation. Synaptic strength of a synapse is strongly dependent on the amount of neurotransmitter (glutamate)

released presynaptically and on the amount of glutamate-responsive a-arnino-3-hydroxy-5-methyl-4-isoxazolepropionic

acid (AMP A) receptors present in the postsynaptic cell membrane. On different stimuli, including calcium influx via the

N-methyl-D-aspartate (NMDA) receptor and brain-derived neurotropic factor (BDNF) binding to pre- and postsynaptic

TRKB receptors, ERI<: and MTOR pathways change synaptic strength by both modulating neurotransmitter (glutamate)

release and by regulating the insertion of AMP A receptors, which are activated by glutamate.

Abbreviations: glut: glutamate, G-protein: guanine nucleotide binding protein, GRF: growth factor, GTP: guanosine

triphosphate, TRF: transcription factor.

5 4 ONCOGENES ON l>IY MIND: ERK AND MTOR SIGN,'\.LING IN COGNITIVE DISEASES

:; :I; .~

""" ;:! ;;; IV "'!1 .....

~ "t

(J1 ('b

(J1 N

ffily1A/\~ W\Vv~~

Translation

I I I I I I I I I I I I I

----

[~:J>roHfer~ti~n, -~~ff~~en~iati~n,Cell s~r-~iv~(]

A) Non-neuronal cell

Presynaptic side

~----+ rERK\ 1/ ~

-- -t,!,eurotransm itter ''It release

s 0 -------...... // l!

---~.,...,',

Postsynaptic side

.... ~ ~

GTP • --~ I

+

I I I I I I I I I I

.. Synaptic pfasticity

~<nl:~~n I 1 receptors

Translation - ..

I I I I I I I I I I I I I I

./ @ranscription

B) Excitatory neuron

References

1. Tartaglia M, Mehler EL, Goldberg R, Zampino G, Brunner HG, Kremer H, et al. Mutations in PTPN11, encoding

the protein tyrosine phosphatase SHP-2, cause Noonan syndrome. Nat Genet 2001;29:465-8.

2. Aoki Y, Niihori T, Kawame H, Kurosawa K, Ohashi H, Tanaka Y, et al. Germline mutations in HRAS proto­

oncogene cause Costello syndrome. Nat Genet 2005;37:1038-40.

3. Niihori T, Aoki Y, Narumi Y, Neri G, Cave H, Verloes A, et al. Germline KRAS and BRAF mutations in cardio­

facial-cutaneous syndrome. Nat Genet 2006;38:294-6.

4. Schubbert S, Zenker M, Rowe SL, Boll S, Klein C, Bollag G, et al. Germline KRAS mutations cause Noonan

syndrome. Nat Genet 2006;38:331-6.

5. Rodriguez-Viciana P, Tetsu 0, Tidyman WE, Estep AL, Conger BA, Cruz MS, et al. Germline mutations in genes

within the MAPK pathway cause cardio-facial-cutaneous syndrome. Science 2006;311:1287-90.

6. Tartaglia M, Pennacchio LA, Zhao C, Yadav KK, Fodale V, Sarkozy A, et al. Gain-of-function SOSl mutations

cause a distinctive form of Noonan syndrome. Nat Genet 2007;39:75-9.

7. Roberts AE, Araki T, Swanson KD, Montgomery KT, Schiripo TA, Joshi VA, et al. Germline gain-of-function

mutations in SOSl cause Noonan syndrome. Nat Genet 2007;39:70-4.

8. Pandit B, Sarkozy A, Pennacchio LA, Carta C, Oishi K, Martinelli S, et al. Gain-of-function RAFl mutations cause

Noonan and LEOPARD syndromes with hypertrophic cardiomyopathy. Nat Genet 2007;39:1007 -12.

9. Razzaque MA, Nishizawa T, Komoike Y, Yagi H, Furutani M, Amo R, et al. Germline gain-of-function mutations in

RAFl cause Noonan syndrome. Nat Genet 2007;39:1013-7.

10. Carta C, Pantaleoni F, Bocchinfuso G, Stella L, Vasta I, Sarkozy A, et al. Germline missense mutations affecting

KRAS Isoform B are associated with a severe Noonan syndrome phenotype. Am J Hum Genet 2006;79:129-35.

11. Brems H, Chmara M, Sahbatou M, Denayer E, Taniguchi K, Kato R, et al. Germline loss-of-function mutations in

SPREDl cause a neurofibromatosis 1-like phenotype. Nat Genet 2007;39:1120-6.

12. van der Burgt I, Thoonen G, Roosenboom N, Assman-Hulsmans C, Gabreels F, Otten B, et al. Patterns of cognitive

functioning in school-aged children with Noonan syndrome associated with variability in phenotypic expression. J

Pediatr 1999;135:707-13.

13. van Engelen SJ, !<:tab LC, Moll HA, de Goede-Bolder A, Pluijm SM, Catsman-Berrevoets CE, et al. Quantitative

differentiation between healthy and disordered brain matter in patients with neurofibromatosis type I using diffusion

tensor imaging. AJNR Am J Neuroradiol2008;29:816-22.

14. Hyman SL, Shores A, North KN. The nature and frequency of cognitive deficits in children with neurofibromatosis

type 1. Neurology 2005;65:1037-44.

15. Levine TM, Materek A, Abel J, O'Donnell M, Cutting LE. Cognitive proffie of neurofibromatosis type 1. Semin

Pediatr Neurol2006;13:8-20.

16. Tyler 'iXlj, Perrett SP, Pozzo-Miller LD. The role of neurotrophins in neurotransmitter release. Neuroscientist

2002;8:524-31.

17. Kushner SA, Elgersma Y, Murphy GG, Jaarsma D, van Woerden GM, Hojjati MR, et al. Modulation of presynaptic

plasticity and learning by the H-ras/extracellular signal-regulated kinase/synapsin I signaling pathway. J Neurosci

2005;25:9721-34.

18. Olijslagers JE, de Kloet ER, Elgersma Y, van Woerden GM, Joels M, Karst H. Rapid changes in hippocampal CAl

pyramidal cell function via pre- as well as postsynaptic membrane mineralocorticoid receptors. Eur J Neurosci

2008;27:2542-50.

19. Costa RM, Federov NB, KoganJH, Murphy GG, Stern], Ohno M, et al. Mechanism for the learning deficits in a

mouse model of neurofibromatosis type 1. Nature 2002;415:526-30.

5 6 ONCOGENES ON MY MIND: .ERK AND MTOR SIGNALING IN COGNITIVE DISEASES

20. Farnsworth CL, Freshney NW, Rosen LB, Ghosh A, Greenberg ME, Feig LA. Calcium activation of Ras mediated

by neuronal exchange factor Ras-GRF. Nature 1995;376:524-7.

21. Krapivinsky G, Krapivinsl.-y L, Manasian Y, Ivanov A, Tyzio R, Pellegrino C, et a!. The NMDA receptor is coupled

to the ERIC pathway by a direct interaction between NR2B and RasGRF1. Neuron 2003;40:775-84.

22. Brarnbilla R, Gnesutta N, Minichiello L, White G, Roylance AJ, Herron CE, et a!. A role for the Ras signalling

pathway in synaptic transmission and long-term memory. Nature 1997;390:281-6.

23. Li S, Tian X, Hartley DM, Feig LA. Distinct roles for Ras-guanine nucleotide-releasing factor 1 (Ras-GRF1) and Ras­

GRF2 in the induction oflong-term potentiation and long-term depression. J Neurosci 2006;26:1721-9.

24. Chen HJ, Rojas-Soto M, Oguni A, Kennedy MB. A synaptic Ras-GTPase activating protein (p135 SynGAP)

inhibited by CaM ltinase II. Neuron 1998;20:895-904.

25. Gelinas JN, Banko JL, Hou L, Sonenberg N, Weeber EJ, Klann E, eta!. ERK and mTOR signaling couple beta­

adrenergic receptors to translation initiation machinery to gate induction of protein synthesis-dependent long-term

potentiation. J Bioi Chern 2007;282:27527-35.

26. Thomas GM, Huganir RL. MAPK cascade signalling and synaptic plasticity. Nat Rev Neurosci 2004;5:173-83.

27. Waltereit R, Weller M. Signaling from cAMP /PKA to MAPK and synaptic plasticity. Mol Neurobiol2003;27:99-106.

28. Zhu JJ, Qin Y, Zhao M, Van Aelst L, Malinow R. Ras and Rap control AMP A receptor trafficlting during synaptic

plasticity. Cell2002;110:443-55.

29. Lee SH, Lee JW, Soung YH, Kim HS, Park WS, Kim SY, et ai. BRAF and KRAS mutations in stomach cancer.

Oncogene 2003;22:6942-5.

30. Ying SW, Futter M, Rosenblum K, Webber MJ, Hunt SP, Bliss TV, et a!. Brain-derived neurotrophic factor induces

long-term potentiation in intact adult hippocampus: requirement for ERK activation coupled to CREB and

upregulation of Arc synthesis. J Neurosci 2002;22:1532-40.

31. Kelleher RJ, 3rd, Govindarajan A, Tonegawa S. Translational regulatory mechanisms in persistent forms of synaptic

plasticity. Neuron 2004;44:59-73.

32. Prather P, de Vries PJ. Behavioral and cognitive aspects of tuberous sclerosis complex. J Child Neurol2004;19:666-

74.

33. Messina MP, Rauktys A, Lee L, Dabora SL. Tuberous sclerosis preclinical studies: timing of treatment, combination

of a raparnycin analog (CCI-779) and interferon-gamma, and comparison of raparnycin to CCI-779. BMC Pharmacal

2007;7:14.

34. Habib SL, Dania! E, Nath S, Schneider J, Jenltinson CP, Duggirala R, eta!. Genetic polymorphisms in OGG1 and

their association with angiomyolipoma, a benign kidney tumor in patients with tuberous sclerosis. Cancer Bioi Ther

2008;7:23-7.

35. Tohda C, Nakanishi R, Kadowaki M. Learning deficits and agenesis of synapses and myelinated axons in

phosphoinositide-3ltinase-deficient mice. Neurosignals 2006;15:293-306.

36. Kwon CH, Luikart BW, Powell CM, Zhou J, Matheny SA, Zhang W, eta!. Pten regulates neuronal arborization and

social interaction in mice. Neuron 2006;50:3 77-88.

37. Fraser MM, Bayazitov IT, Zaltharenko SS, Baker SJ. Phosphatase and tensin homolog, deleted on chromosome 10

deficiency in brain causes defects in synaptic structure, transmission and plasticity, and myelination abnormalities.

Neuroscience 2008;151:476-88.

38. Goorden SM, vao Woerden GM, van der Weerd L, CheadleJP, Elgersma Y. Cognitive deficits in Tscl+/- mice in

the absence of cerebral lesions and seizures. Ann Neural 2007;62:648-55.

39. Ehninger D, HanS, Shilyansky C, Zhou Y, Li W, Kwiatkowski DJ, eta!. Reversal of learning deficits in a Tsc2+ /­

mouse model of tuberous sclerosis. Nat Med 2008;14:843-8.

Cr! \PTER 2 5 7

40. Bekinschtein P, Katche C, Slipczuk LN, Igaz LM, Cammarota M, Izquierdo I, et al. mTOR signaling in the

hippocampus is necessary for memory formation. Neurobiol Learn Mem 2007;87:303-7.

41. Tang SJ, Reis G, Kang H, Gingras AC, Sonenberg N, Schuman EIYL A rapamycin-sensitive signaling pathway

contributes to long-term synaptic plasticity in the hippocampus. Proc Nad Acad Sci US A 2002;99:467-72.

42. Takei N, Inamura N, Kawamura M, Namba H, Hara K, Yonezawa K, et al. Brain-derived neurotrophic factor

induces mammalian target of rapamycin-dependent local activation of translation machinery and protein synthesis in

neuronal dendrites. J Neurosci 2004;24:9760-9.

43. Gong R, Park CS, Abbassi NR, Tang SJ. Roles of glutamate receptors and the mammalian target of rapamycin

(mTOR) signaling pathway in activity-dependent dendritic protein synthesis in hippocampal neurons. J Bioi Chern

2006;281 :18802-15.

44. Pfeiffer BE, Huber KM. Current advances in local protein synthesis and synaptic plasticity. J Neurosci 2006;26:7147-

50.

45. Raab-Graham KF, Haddick PC, Jan YN, Jan LY. Activity- and roTOR-dependent suppression of Kvl.l channel

mRNA translation in dendrites. Science 2006;314:144-8.

46. Man HY, Wang Q, Lu WY, Ju W, Ahmadian G, Liu L, et al. Activation of PI3-kinase is required for AMP A receptor

insertion during LTP of mEPSCs in cultured hippocampal neurons. Neuron 2003;38:611-24.

47. Williams PG, Hersh JH. Brief report: the association of neurofibromatosis type 1 and autism. J Autism Dev Disord

1998;28:567 -71.

48. Zafeitiou DI, Ververi A, Vargiami E. Childhood autism and associated comorbidities. Brain Dev 2007;29:257-72.

49. Ferner RE, Huson SM, Thomas N, Moss C, Willshaw H, Evans DG, et al. Guidelines for the diagnosis and

management of individuals with neurofibromatosis 1. J Med Genet 2007;44:81-8.

50. Li W, Cui Y, Kushoer SA, Brown RA, Jentsch JD, Frankland PW, et al. The HMG-CoA reductase inhibitor

lovastatin reverses the learning and attention deficits in a mouse model of neurofibromatosis type 1. Curr Bioi

2005;15:1961-7.

51. Zeng LH, Xu L, Gutmann DH, Wong M. Rapamycin prevents epilepsy in a mouse model of tuberous sclerosis

comple.'<. Ann Neurol2008;63:444-53.

52. Denayer E, Legius E. What's new in the neuro-cardio-facial-cutaneous syndromes? Eur J Pediatr 2007;166:1091-8.

53. Konstantinopoulos P A, Karamouzis MV, Papavassiliou AG. Post-translational modifications and regulation of the

RAS superfamily of GTPases as anticancer targets. Nat Rev Drug Discov 2007;6:541-55.

54. Vase L.R, Arora G.S., O'Brien PF, Hannan F. Pharmacologic Rescue of Behavioral Deficits in Drosophila NF1

Mutants. NF Conference (abstract) 2008:p. 97.

55. Graaf MR, Riche! DJ, van Noorden CJ, Guchelaar HJ. Effects of statins and farnesyltransferase inhibitors on the

development and progression of cancer. Cancer Treat Rev 2004;30:609-41.

56. Belelli D, Lambert JJ. Neurosteroids: endogenous regulators of the GABA(A) receptor. Nat Rev Neurosci

2005;6:565-75.

57. Kohno M, Pouyssegur]. Targeting the ERK signaling pathway in cancer therapy. Ann Med 2006;38:200-11.

58. Franz DN, Leonard J, Tudor C, Chuck G, Care M, Sethurarnan G, et al. Rapamycin causes regression of

astrocytomas in tuberous sclerosis complex. Ann Neurol 2006;59:490-8.

59. Bissler JJ, McCormack FX, Young LR, Elwing JM, Chuck G, Leonard JM, et al. Sirolimus for angiomyolipoma in

tuberous sclerosis complex or lymphangioleiomyomatosis. N Eng!J Med 2008;358:140-51.

60. Cloughesy TF, Yoshimoto K, Nghiemphu P, Brown K, Dang J, Zhu S, et al. Antitumor activity of rapamycin in a

Phase I trial for patients with recurrent PTEN-deficient glioblastoma. PLoS Med 2008;5:e8.

61. Buerger C, DeVries B, Stambolic V. Localization of Rheb to the endomembrane is critical for its signaling function.

Biochem Biophys Res Commun 2006;344:869-80.

5 8 ONCOGENES ON JV!Y MIND: ERKAND MTOR SIGNALING IN COGNITIVE DISEASES

62. Zenker M, Lehmann K, Schulz AL, Barth H, Hansmann D, Koenig R, et al. Expansion of the genotypic and

phenotypic spectrum in patients with ERAS germline mutations. J Med Genet 2007;44:131-5.

63. Nava C, Hanna N, Michot C, Pereira S, Pouvreau N, Niihori T, et al. Cardia-facial-cutaneous and Noonan

syndromes due to mutations in the RAS/MAPK signalling pathway: genotype-phenotype relationships and overlap

with Costello syndrome. J Med Genet 2007;44:763-71.

64. Ktab LC, de Goede-Bolder A, Aarsen FK, Pluijm SM, Bouman MJ, van der Geest JN, et al. Effect of simvastatin on

cognitive functioning in children with neurofibromatosis type 1: a randomized controlled trial. J ama 2008;300:287-

94.

65. Ahles TA, Saykin AJ. Candidate mechanisms for chemotherapy-induced cognitive changes. Nat Rev Cancer

2007;7:192-201.

66. Vardy J, Wefel JS, Ahles T, Tannock IF, Schagen SB. Cancer and cancer-therapy related cognitive dysfunction: an

international perspective from the Venice cognitive workshop. Ann Oncol2008;19:623-9.

67. Jenkins V, Shilling V, Deutsch G, Bloomfield D, Morris R, Allan S, et al. A 3-year prospective study of the effects of

adjuvant treatments on cognition in women with early stage breast cancer. Br J Cancer 2006;94:828-34.

68. Atkins CM, Selcher JC, Petraitis JJ, Trzaskos JM, Sweatt JD. The MAPK cascade is required for mammalian

associative learning. Nat Neurosci 1998;1:602-9.

69. Lee DA, Portnoy S, Hill P, Gillberg C, Patton MA. Psychological profile of children with Noonan syndrome. Dev

Med Child Neurol2005;47:35-8.

70. Shaw AC, Kalidas K, Crosby AH, JefferyS, Patton MA. The natural history of Noonan syndrome: a long-term

follow-up study. Arch Dis Child 2007;92:128-32.

71. Edouard T, Montagner A, Dance M, Conte F, YarrA, Parfait B, et al. How do Shp2 mutations that oppositely

influence its biochemical activity result in syndromes with overlapping symptoms? Cell Mol Life Sci 2007;64:1585-90.

72. Kerr B, Delrue MA, Sigaudy S, Pcrvccn R, Marche M, Burgelin I, et al. Genotype-phenotype correlation in Costello

syndrome: HRAS mutation analysis in 43 cases. J Med Genet 2006;43:401-5.

73. Gripp KW, Lin AE, Stabley DL, Nicholson L, Scott CI, Jr., Doyle D, et al. HRAS mutation analysis in Costello

syndrome: genotype and phenotype correlation. Am J Med Genet A 2006;140:1-7.

74. Estep AL, Tidyman WE, Teitell MA, Cotter PD, Rauen KA. HRAS mutations in Costello syndrome: detection of

constitutional activating mutations in codon 12 and 13 and loss of wild-type allele in malignancy. Am J Med Genet A

2006;140:8-16.

75. Axelrad ME, Nicholson L, Stabley DL, Sol-Church K, Gripp KW. Longitudinal assessment of cognitive

characteristics in Costello syndrome. Am J Med Genet A 2007;143:3185-93.

76. Delrue l'vlA, Chateil JF, Arveiler B, Lacombe D. Costello syndrome and neurological abnormalities. Am J Med Genet

A 2003;123:301-5.

77. Narumi Y, Aoki Y, Niihori T, Neri G, Cave H, Verloes A, et al. Molecular and clinical characterization of cardia­

facial-cutaneous (CFC) syndrome: overlapping clinical manifestations with Costello syndrome. Am J Med Genet A

2007;143:799-807.

78. Yoon G, Rosenberg J, Blaser S, Rauen KA. Neurological complications of cardia-facial-cutaneous syndrome. Dev

Med Child Neurol2007;49:894-9.

79. Armour CM, Allanson JE. Further delineation of cardia-facial-cutaneous syndrome: clinical features of 38 individuals

with proven mutations. J Med Genet 2008;45:249-54.

80. Jansen AC, Sancak 0, D'Agostino MD, Badhwar A, Roberts P, Gobbi G, et al. Unusually mild tuberous sclerosis

phenotype is associated with TSC2 R905Q mutation. Ann Neurol2006;60:528-39.

81. Luat AF, Makki M, Chugani HT. Neuroimaging in tuberous sclerosis complex. Curr Opin Neurol2007;20:142-50.

CH,\l'TFR 2 5 9

82. Lachlan KL, Lucassen AM, Bunyan D, Temple IK Cowden syndrome and Bannayan Riley Ruvalcaba syndrome

represent one condition with variable expression and age-related penetrance: results of a clinical study of PTEN

mutation carriers. J Med Genet 2007;44:579-85.

83. Eng C. PTEN: one gene, many syndromes. Hum Mutat 2003;22:183-98.

84. Hendriks YM, Verhallen JT, van der Smagr JJ, Kant SG, Hilhorst Y, Hoefsloot L, et al. Bannayan-Riley-Ruvalcaba

syndrome: further delineation of the phenotype and management of PTEN mutation-positive cases. Fam Cancer

2003;2:79-85.

85. Lok C, Viseux V, Avril MF, Richard MA, Gondry-Jouet C, Deramond H, et al. Brain magnetic resonance imaging in

patients with Cowden syndrome. Medicine (Baltimore) 2005;84:129-36.

86. Giese KP, Friedman E, Telliez JB, Fedorov NB, Wines M, Feig LA, et al. Hippocampus-dependent learning and

memory is impaired in mice lacking the Ras-guanine-nucleotide releasing factor 1 (Ras-GRF1). Neuropharmacology

2001;41:791-800.

87. Femandez-Medarde A., Esteban LM, Nunez A, Porteros A, Tessarollo L, Santos E. Targeted disruption of Ras-Gr£2

shows its dispensability for mouse growth and development. Mol Cell Bioi 2002;22:2498-504.

88. Kamiyama NH, Watabe AM, Carlisle HJ, Porter K, Charlesworth P, Monti], et al. SynGAP regulates ERK/MAPK

signaling, synaptic plasticity, and learning in the complex with postsynaptic density 95 and NMDA receptor. J

Neurosci 2002;22:9721-32.

89. von der Brelie C, Waltereit R, Zhang L, Beck H, Kirschstein T. Impaired synaptic plasticity in a rat model of

tuberous sclerosis. Eur J Neurosci 2006;23:686-92.

90. Mana beT, Aiba A, Yamada A, Ichise T, Sakagami H, Kondo H, et al. Regulation of long-term potentiation by H-Ras

through NMDA receptor phosphorylation. J Neurosci 2000;20:2504-11.

91. Guilding C, McNair K, Stone TW, Morris BJ. Restored plasticity in a mouse model of neurofibromatosis type 1 via

inhibition of hyperactive ERK and CREB. Eur J Neurosci 2007;25:99-105.

92. Costa RM, Silva AJ. Mouse models of neurofibromatosis type I: bridging the GAP. Trends Mol Med 2003;9:19-23.

93. Chen AP, Ohoo M, Giese KP, Kuho R, Chen RL, Silva AJ. Forebrain-specific knockout of B-raf kinase leads to

deficits in hippocampal long-term potentiation, learning, and memory. J Neurosci Res 2006;83:28-38.

94. Shalin SC, Zirtgiebel U, Honsa KJ, Julien JP, Miller FD, Kaplan DR, et al. Neuronal MEK is important for normal

fear conditioning in mice. J Neurosci Res 2004;75:760-70.

95. Kelleher RJ, 3rd, Govindarajan A,Jung HY, Kang H, Tonegawa S. Translational control by MAPK signaling in long­

term synaptic plasticity and memory. Cell2004;116:467 -79.

96. Mazzucchelli C, Vantaggiato C, Ciamei A, Fasano S, Pakhotio P, Krezel W, et al. I<:nockout of ERI<J MAP kinase

enhances synaptic plasticity in the striatum and facilitates striatal-mediated learning and memory. Neuron

2002;34:807 -20.

97. Selcher JC, Nekrasova T, Paylor R, Landreth GE, Sweatt JD. Mice lacking the ERK1 isoform of MAP kinase are

unimpaired in emotional learning. Learn Mem 2001;8:11-9.

98. Satoh Y, Endo S, Ikeda T, Yamada K, Ito M, Kuroki M, et al. Extracellular signal-regulated kinase 2 (ERK2)

knockdown mice show deficits in long-term memory; ERK2 has a specific function in learning and memory. J

Neurosci 2007;27:10765-76.

99. Waltereit R, Welzl H, Dichgans J, Lipp HP, Schmidt WJ, Weller M. Enhanced episodic-like memory and kindling

epilepsy in a rat model of tuberous sclerosis. J Neurochem 2006;96:407 -13.

6 0 ONCOGENES ON MY MIND: ERKAND MTOR SIGNAl.ING IN COGNITIVE DISEASES

CH,\P'rER 2 6 1

Example if the %JI Complex Figure Test (lift), with the copy (middle) and delqyed recall (drawn by heart

after 15 minutes; right) versions drawn by a 1 0-year old girl with NF1.

Lianne C. :&:tab, MSc1•3, Femke K. Aarse:n, MA2,Arja de Goede~ Bolder, MD1,

Co:riene E. Catsma:n~Ber.tevoets, MD PhD2, Willem F. Arts, MD PhD2,

Heruiette A. Moll, MD PhD1, Ype Elgersma, PhD3

The NFJ CoRe Team (Cognitive Research Team)~ Erasmus MC/Sophia Children~s Hospital, Rotterdam, The Netherlands. 1 Department of General Pediatrics, 2 Department of Pediatric Neurology,

3 Departn1ent of Neuroscience.

Journal of Child Neurolog~ in press (2008)

Abstract

School functioning of 86 Dutch Neurofibromatosis type 1 children (7 -17 years) was analyzed

using teacher questionnaires to determine the impact of Neurofibromatosis type 1 on school

performance. In all, 75% of the Neurofibromatosis type 1 children performed more than one

standard deviation below grade peers in at least one of the domains of spelling, mathematics,

technical reading or comprehensive reading. Furthermore, Neurofibromatosis type 1 children

had a 4-fold increased risk for attending special education, and a 6-fold increased risk for

receiving remedial teaching for learning, behavior, speech and/ or motor problems. Children

without any apparent learning disability still frequently displayed neuropsychological deficits.

Only 10% of the children did not show any school-functioning problems. Finally, it was found

that the clinical severity of Neurofibromatosis type 1 correlated with the cognitive deficits.

Taken together, we show that Neurofibromatosis type 1 has profound impact on school

performance. Awareness of these problems may facilitate timely recognition and appropriate

support.

Keywords:

Neurofibromatosis Type 1 / learning disabilities / children

6 4 THE IMPACT OF NEUROFIBROI\iATOSIS Til'E 1 ON SCHOOL PERFORMANCE

Introduction

Neurofibromatosis type 1 (NF1) is an autosomal dominant disease with an incidence of 1 in

2500 to 3000 individuals, of whom 50% are de novo cases .I Characteristics of NF1 are various

neurocutaneous manifestations, including cafe au lait maculae, axillary freckling, neurofibromas,

and Lisch nodules. Minor disease characteristics include developmental delay, poor motor skills

and speech problems.2 Clinical presentation of NF1 is highly variable even within families. In

children, the most frequent complication of NF1 1s a cognitive impairment.3

Neuropsychological deficits include a lowered average intelligence quotient (IQ), problems with

visual-spatial skills, memory, language, executive functioning, and attention (reviewed in 4). In

addition, children with NF1 have poor social skills, and up to 40% have attention-deficit

hyperactivity disorder. s. 6

Until recendy, litde attention was paid to the impact of NF1 on school performance. Estimates

for the occurrence of learning disabilities vary from 30 to 70%.6-B This large variation is likely

due to different definitions for learning disabilities, selection bias of the study groups, or small

sample size (reviewed in 4). Importandy, all previous studies on learning disabilities associated

with NF1 have solely used academic achievement tests taken at a clinical setting. Although these

measurements reliably reflect the academic achievement level of a child, they do not take into

account how this level was achieved. Receiving intensive remedial teaching or special education

may seriously confound the interpretation of academic achievement test scores, and this may

result in a significant underestimation of the learning disabilities and school problems associated

with NFL Therefore, to get a more realistic assessment of school performance, it is important

to combine these different types of information on school functioning. However, quantitative

studies on the level of special education, remedial teaching, or grade repetition in these children

are largely absent.

The aim of this study is to determine the impact of NF1 on school performance of Dutch

children with NF1 by examining the learning disabilities displayed in the school setting and by

determining the relative risk for receiving extra support in the form of grade repetition, remedial

teaching, and special education. In addition, we investigated the relationship between the clinical

severity of NF1 and the cognitive deficits.

CH.\l'J'ER 3 6 5

Methods

Patients

Participants were recruited from the multidisciplinary pediatric NF1 outpatient clinic of the

Erasmus MC - Sophia Children's Hospital, Rotterdam. This outpatient clinic is a supraregional

reference center that predominantly receives patients from the southwestern part of the

Netherlands (about 3 million citizens). Data for this study were obtained in the context of a

larger ongoing study on NF1 and cognitive functioning. Inclusion criteria were: NF1 diagnosis

according to the criteria of the National Institutes of Health,9 7 to 17 years of age for the school

performance questionnaire, 8 to 17 years of age for the neuropsychological assessment, and

informed consent from parents and children aged older than 12 years. Exclusion criteria were:

segmental NF1, pathology of the central nervous system (other than asymptomatic gliomas),

deafness, severely impaired vision, use of anti-epileptics, inefficient production or

comprehension of the Dutch language, and severe mental retardation (Full Scale IQ below 48,

to exclude children with an IQ score below the range covered by the Wechsler Intelligence Scale

for Children, Revised, Dutch version).

In all, 126 children fulfilled age criteria. Twelve children were excluded on the basis of

segmental NF1 (n=3), use of anti-epileptics (n=3), pathology of the central nervous system

(hydrocephalus, n=3), severe mental retardation (n=1) and inefficient production or

comprehension of the Dutch language (n=2). The remaining 114 children were invited to

participate in the school performance questionnaire and neuropsychological assessment.

Disease severity was scored by an experienced pediatrician of the NF1 team (A. G.B.) according

to the Riccardi Scale,lO modified to exclude cognitive aspects of NFL Minimal NF1 was scored

in the absence of features that compromise health (when only harmless clinical features such as

cafe au lait maculae, freckling, and Lisch nodules were present). Mild NF1 was scored when

minor medical complications, such as small stature or discrete plexiform neurofibroma, were

present. Moderate NF1 was scored in case of complications that were a significant compromise

to health, such as paravertebral neurofibromas or hypertension. Severe NF1 was scored in case

of malignancy. Familial or sporadic NF1 was determined by the pediatrician from the family

history. Informed consent was received from all participants. This study was approved by the

medical ethical committee of the Erasmus MC- Sophia Children's Hospital.

6 6 THE IMPACT OF NELTROFlllRO!vLUOSIS TVPE 1 ON SCHOOL PERFOlli\IANCE

School Performance

Teachers of the participating patients were requested to complete an abbreviated version of the

Teacher's Report Form,11 with additional quantitative and qualitative questions on remedial

teaching. Teachers reported the most recent scores on technical reading, comprehensive

reading, spelling, and mathematics from the Dutch Student Monitoring System. This system is a

government-enforced system of standardized and validated academic performance tests for

technical reading, comprehensive reading, spelling, and mathematics assessed 3 times a year in

all grades of primary school in the Netherlands.12 The tests most frequently used for technical

reading, comprehensive reading, spelling, and mathematics are all rated sufficient to good on

norms, good on reliability, and good on construct validity by the Committee on Test Affairs

Netherlands. Key cutoff criterion for admission to a school for the mentally retarded in the

Netherlands is IQ<60. Cutoff criteria for admission to a school for the learning disabled are 1)

IQ 61-89, or 2) IQ>SO with a learning efficacy below 75%, or 3) IQ in the normal range in

combination with severe visual, speech- language and hearing, motor, social-emotional, or

behavioral problems_13. 14

Neuropsychological Assessment

Neuropsychological tests developed for children were administered to assess cognitive skills in

six domains: 1. Intelligence (Wechsler Intelligence Scale for Children, Revised, Dutch version),

2. Memory (Rey Auditory Verbal Learning Test for verbal memory; Rey Complex Figure Test­

delayed recall for nonverbal memory), 3. Language (Peabody Picture Vocabulary Test III for

receptive language; Boston Naming Test for expressive language), 4. Visual-spatial skills

Gudgment of Line Orientation task for line orientation; Rey Complex Figure Test - copy for

visual integration; and Beery Developmental Test of Visual-Motor Integration for visual motor

integration), 5. Executive skills (Trailmaking Test A and B for rote memory and divided

attention; Animal naming for verbal fluency; Wisconsin Card Sorting Test for concept

formation and perseverations), and 6. Attention (Stroop Color-Word Test for Selective

attention; Cancellation Test - speed for sustained attention; Cancellation Test - attention

fluctuations for attention fluctuations).1 5• 16 All tests were administered in their Dutch versions

and scored by a single pediatric neuropsychologist. To allow comparison across ages,

neuropsychological scores were converted into Z-scores (deviation from the mean of a normal

population). The evaluator was not informed on the medical status or school results of the

patient. The neuropsychological testing took 3.5 hours to complete, and was divided into 2

sessions (2 hours and 1.5 hours) with a break of 1 hour in between.

Cii\PThR:l 6 7

Definitions

REMEDIAL TEACHING: Structural assistance on top of regular class assistance offered for

problems with learning, motor function, speech, and/ or behavior.

DIDACTIC SCORE: The test-specific score on one of the didactic tests used in the student

monitoring system.

LEARNING EFFICACIES: To compare students across tests and ages, didactic scores are converted

into learning efficacies. Hereto, the didactic score was first converted into a didactic age

equivalent using normative conversion tables.12 Learning efficacy was then calculated by dividing

the determined didactic age equivalent by the actual months of education a child received in

primary school(= didactic age). One school year consists of 10 didactic months. For example, if

a child at the end of 5th grade (didactic age 5 x 10 =50 months) has a didactic age equivalent on

mathematics of 40 months, the learning efficacy of this child is 40/50 x 100%= 80%. Didactic

age equivalents are required by the Dutch government for reporting progress of students.17 The

normative average score at a certain didactic age per definition equals a learning efficacy of

100% (didactic age= didactic age equivalent).

LEARNING DISABIUTY: Learning disability was defined as a learning efficacy for technical

reading, comprehensive reading, spelling, or mathematics of more than 1 standard deviation

below the average (< 85%). A learning disability was termed specific if occurring with a normal

IQ (lQ ?:: 85), and termed general with an IQ of more than 1 standard deviation below the

mean (lQ < 85). To determine whether there was a specific or general learning disability, IQ

was obtained from neuropsychological assessment or from the school performance

questionnaire (question X of the Teacher's Report Form if the reported score was less than 1

year old and obtained with Wechsler Intelligence Scale for Children, version III or Revised,

Dutch version). When learning efficacies for all four didactic domains were available, patients

were assigned to one of the following groups: a No Learning Disabilities group, a Specific

Learning Disabilities group (children with learning disabilities on one or more of the didactic

domains but a normal IQ) or a General Learning Disabilities group (children with learning

disabilities on one or more of the didactic domains and an IQ of more than 1 SD below the

mean). If for a particular domain, scores were not present or if only qualitative instead of

quantitative scores were available, these scores were treated as missing values in that specific

domain.

Statistical Analysis

Data were analyzed in SPSS 12.0 (SPSS Inc, Chicago, Illinois) using a parametric test for

continuous variables (2-sided independent t-test, Analysis of Variance [ANOVA] with a post-

6 8 THE IMPACT OF NEUROFIBROMATOSIS Til'E 1 ON SCHOOL PERFORJ\1ANCE

hoc 2-sided t-test) and nonparametric tests for categorical variables or if n<20 (binomial test,

chi-square test, Kruskal-Wallis test with post hoc Mann-Whitney test). The chi-square test was

used to compare variance of our study group with that of the normal population. The

Kolmogorov-Smirnov test was used to control for normal distribution.

Results

Informed consent for the school performance questionnaire was received for 89 children

(response 78%). The questionnaire was completed by 86 of the 89 teachers (response 97%).

School type was specified for all 86 children. Learning efficacies for one or more of the domains

of technical reading, comprehensive reading, spelling, or mathematics and presence or absence

of a learning disability could be calculated from quantitative scores of 75 children. These

included 70 scores for technical reading, 61 for comprehensive reading, 69 for spelling, and 65

for mathematics. For 54 children, all 4 didactic scores were available. Remedial teaching was

scored for 75 children and grade repetition for 70 children. Informed consent for the

neuropsychological examination was received for 62 children (response 54%). Table 1 provides

an overview of the patient characteristics.

Table 1: Patient characteristics

Characteristic

Sex: Male/Female

Age at assessment (years)

Familial NFl/Sporadic NF1/unconfumed

Modified Riccardi scale

Scale I (minimal)

Scale IT (nilld)

Scale Ill (moderate)

Scale N (severe)

Using medication for attention deficit-hyperactivity disorder

NFl: Neurofibromatosis type 1.

Learning Disabilities

Number of patients (n-86)

47/39 (54.7 /453%)

11.9 ± 2.5

36/48/2

30

30

25

14

School performance of children with NF1 was substantially impaired on all 4 domains of

technical reading, comprehensive reading, spelling, and mathematics (table 2). Mean learning

efficacies were significantly lower than normative grade-peer average (100%). Children with

NF1 follow an average learning efficacy curve of 7 5% (at -1.7 SD from average, t=-7 .66,

U!\PTER3 6 9

p<O.OOOS), which is at the cutoff level of admission to a school for the learning disabled.

Learning disabilities (learning efficacy below 85% [<-1 SD]) were present in at least 56/75

(75%) of the children with NFL Even when using a -2 SD cut-off (learning efficacy < 70%),

47/75 children (63%) displayed impairments in one or more didactic domains. Using IQ data of

these children, we were able to determine whether a child had a specific learning disability for a

given didactic domain (a learning efficacy on a specific didactic domain below 85% but with

normal IQ) or a general learning disability (a learning efficacy on a specific didactic domain

below 85% '.Vi.th IQ<85). This showed that specific and general learning disabilities were

distributed equally over the academic areas (see table 2).

Table 2: Overview of the learning disabilities found in our patient group per didactic domain.

Mean learning Specific learning General Learning

Didactic domain efficacy in % (SD) disability (%) disability(%)

Spelling 70 (31)** 19/69 (28%) 26/69 (38%)

Technical Reading 74 (30)** 19/70 (27%) 25/70 (36%)

Mathematics 77 (35)** 15/65 (23%) 25/65 (39%)

Comprehensive Reading 78 (35)** 12/61 (20%) 22/61 (36%)

**: P-value <0.0005 compared to normative grade peer average (100%).

For 54 children, quantitative school performance data on all 4 didactic domains were present.

This allowed us to assign them to a General Learning Disabilities group (learning disabilities on

one or more of the didactic domains, and IQ<85), a Specific Learning Disabilities group

(learning disabilities on one or more of the didactic domains, but IQ2:85) or to a No Learning

Disabilities group (children without learning disabilities in any of the four didactic domains). On

the basis of these criteria, 21 children (39%) were assigned to the General Learning Disabilities

group, 21 children (39%) to the Specific Learning Disabilities group and only 12 children (22%)

to the No Learning Disabilities group.

Remedial Teaching, Special Education and Grade Repetition

Special education was attended by 37% of the children, which is an odds ratio of 4.1 compared

with the average populationls (Table 3). In total, 33% attended a school for the learning

disabled, and 5% attended a school for the mentally retarded. Forty percent of the children with

NF1 repeated a grade in their school career. In primary school, this was significandy more

frequent compared with the regular population (17% versus 1.9%, binomial test, p<0.0005).19

The majority (85%) of the children with NF1 received remedial teaching for learning problems

(didactical remedial teaching), fine and gross motor problems (motorical), speech problems

(logopedical) and/ or behavioral problems (behavioral), an odds ratio of 5.6 compared with 15%

7 0 THE IMPACT OF NEUROFIBROIVL>\TOSIS TYPE 1 ON SCHOOL PERFORl\tANCE

in the Dutch population.zo A combination of 2 or more types of remedial teaching was given to

52% of all children, mostly including remedial teaching for learning problems.

Nineteen out of 75 children had no learning disabilities. However, 12 (63%) of these children

received remedial teaching, of which 5 (40%) specifically for learning problems. If we look at

the children who were scored for the school type, grade repetition, remedial teaching, and

learning disabilities (n=61), only 6 children (10%) did not have problems on either of these 4

aspects of school performance.

Table 3: Impact of NF1 on school performance.

Type of education (n-86)t

- Special education

- School for learning disabled

- School for mentally retarded

- Regular education

Repeated grade (n=70)t

- in kindergarten

- in primary school

- in secondary school

Remedial teaching given (n=75)§

-Yes

- In primary school (n =52)

-No

Type of remedial teaching (n=64)

-Didactical

-Motorical

-Logopedical

-Behavioral

-Not specified

Patient group

o/o(n)

37% (32)

33% (28)

4.7% (4)

63% (54)

40% (28)

19% (13)

17% (12)

4% (3)

85% (64)

85% (44)

15% (11)

73% (47)

42% (27)

36% (23)

22% (14)

3%(2)

Dutch population

%

9.0%

8.3%

0.6%

91.0%

14.3%

1.9%

15%

85%

Odds Ratio for NF1

(95% confidence interval)

4.1 (3.1-5.4)**

3.9 (2.9-5.3)**

73 (2.8-19.0)**

0.7 (0.6-0.8)**

1.3

9.0**

5.6 (5.1-6.3)**

0.2 (0.1-0.2)**

** P-value binomial test < 0.0005. tReference values from Dutch Ministry of Education, Culture and Science (N -

2.597.700)18• tReference values from Dutch Inspection of Education19, Confidence interval not available. §Reference

values from Matthijsen eta!. (children 8-11 years, N=9.734)20. NF1: Neurofibromatosis type 1.

Neuropsychological assessment

Out of the 86 participating children, 62 (70%) consented to a neuropsychological assessment.

These children showed a mean Full Scale IQ of 86, which was distributed normally. Mental

retardation (Full Scale IQ<70) was present in 11 children (18%). The performance IQ proflle

showed a clip in the scores for block design and object assembly. Compared with normative

Table 4: Neuropsychological results in learning disability groups.

TotalNF1 NoLD group, SLD group, GLD group,

group, n=62; n=S; n=16; n=17;

NeuroEs~cholo~cal Test* mean (SD) mean (SD) mean (SD) mean(SD) tt

Intelligencet

Full Scale I Q 86.2 (15.3) 96.8 (12.1) 97.9 (8.1) 72.6 (7.5) .2,3

Verbal IQ 86.7 (16.2) 98.4 (11.4) 98.2 (10.0) 71.4 (8.5) 2,3

Performance IQ 88.7 (14.7) 95.6 (11.6) 98.4 (12.2) 79.9 (8.7) 2.3

Verbal Comprehension Index 88.5 (14.9) 97.1 (8.8) 99.5 (8.4) 76.0 (9.9) 2.3

Perceprual Organisation Index 88.0 (14.7) 94.6 (10.7) 97.5 (11.2) 79. (10.1) 2,3

Freedom from Distractibility 88.4 (16.8) 102.1 (14.4) 97.6 (14.3) 73.8 (7.2) 2,3

Information 7.8 (2.9) 9.3 (3.4) 9.5 (1.8) 6.1 (2.1) 2,3

Similarities 9.0 (3.2) 10.6 (1.9) 11.4 (2.0) 6.3 (2.2) 2,3

Arithmetic 7.3 (3.9) 10.4 (2.0) 8.3 (4.1) 4.1 (1.8) 2,3

Vocabulary 7.4 (2.7) 9.0 (1.7) 9.2 (1.6) 5.4 (1.8) 2,3

Comprehension 7.9 (2.7) 9.4 (1.8) 9.3 (2.3) 6.3 (1.8) 2,3

Digit Span 8.0 (3.4) 10.1 (3.8) 10.5 (2.8) 5.2 (2.1) 2.3

Picture completion 9.3 (3.5) 9.1 (2.5) 12.3 (3.3) 8.0 (2.9) 1.2

Picture arrangement 9.6 (3.0) 10.5 (2.2) 10.5 (3.0) 8.4 (3.2)

Block Design 7.1 (2.6) 8.4 (2.2) 8.6 (2.7) 5.8 (2.1) 2.3

Object assembly 7.1 (3.2) 8.1 (3.3) 7.8 (2.5) 6.6 (2.6)

Coding 9.3 (2.7) 10.5 (2.3) 9.9 (2.6) 8.7 (2.4)

Mazes 8.2 (3.2) 10.1 (2.6) 8.6 (3.2) 6.7 (2.4)

Memory

Rey A VL T - inlmediate recall -0.03 (1.09) 0.04 (1.41) 0.34 (1.01) -0.53 (1.04)

Rey A VLT- delayed recall 0.11 (1.04) 0.18 (1.18) 0.37 (0.85) -0.18 (1.16)

Rey CPT- delayed recall -1.64 (0.99) -1.86 (0.75) -1.48 (0.83) -1.54 (0.72)

Language

PPVT 0.31 (1.11) 0.93 (0.66) 0.86 (0.65) -0.26 (1.02) 2,3

Boston Naming test -1.12 (1.75) -0.44 (0.98) -0.24 (1.15) -1.70 (1.82)

Visual-spatial skills

Judgement of Line Orientation -1.37 (1.45) -1.65 (0.94) -0.56 (1.66) -1.68 (1.13)

Rey Complex Figure Test- copy -1.28 (1.28) -0.36 (0.82) -0.79 (0.96) -1.68 (1.25) 2,3

BeeryVMI -1.16 (0.84) -0.38 (0.66) -0.85 (0.66) -1.53 (0.70) 2,3

Executive skills

Trailmaking Test A -0.69 (1.12) -1.09 (1.24) -0.69 (0.87) -0.78 (1.36)

Trailmaking Test B -0.58 (1.10) -1.24 (1.20) -0.35 (1.19) -0.86 (1.04)

Animal naming -0.08 (1.23) 0.88 (1.48) 0.06 (1.27) -0.67 (0.91)

Wisconsin CST - perseverations -0.19 (1.13) -0.06 (0.50) 0.13 (1.28) -0.84 (1.20)

Wisconsin CST- categories -0.23 (1.18) 0.26 (0.96) 0.02 (1.09) -0.85 (1.34)

Attention

Stroop Color-Word Test- speed -0.35 (1.94) -0.60 (1.24) 0.67 (1.74) -1.15 (1.77)

Cancellation Test- speed -1.02 (1.70) -1.60 (0.70) -0.02 (1.60) -1.67 (1.80) 1,2

Cancellation Test- AF-1: -2.75 (1.49) -3.34 (0.84) -2.17 (1.11) -3.02 (1.92)

Cancellation Test- corrections§ 1.0 (1.6) 0.9 (1.0) 1.4 (2.4) 0.7 (0.8)

Cancellation Test- omissions** 15.5 (16.1) 15.3 (6.3) 12.9 (8.6) 20.6 (24.4)

*Scores are Z-scores (deviation from the mean of a normal population) unless otherwise indicated. tScores are standard

scores (normative Scale and Index scores: mean=100, D=15; subtest scores: mean=10, SD=3). *Standard deviation of

speed in seconds (no Z-score available; normative score for 12 year-olds: mean=1.7 seconds, interquartile range 2.2-1.5).

§Number of corrections (no Z-score available; cutoff for clinical significance: >3 corrections). **number of omissions (no

Z-score available; cutoff for clioical significance: >3 corrections).

72 THE lMPACT OF NEUROFIBROMATOSIS TYPE 1 ON SCHOOL PERFOR.J\L~NCE

tJ-Kruskal-Wallis, post hoc Mann-Whitney p<O.OS for: 1 = Comparison No Learning Disabilities -Specific Learning

disabilities groups; 2 = Comparison Specific Learning Disabilities - General Learning Disabilities groups; 3 = Comparison

No Learning Disabilities- General Learning Disabilities groups.

NF1: Neurofibromatosis type 1, SD: Standard deviation, SLD: Specific Learning Disabilities, GLD: General Learning

Disabilities, NoLD: No Learning Disabilities, Rey AVLT: Rey Auditory Verbal Learning Test, Rey CPT: Rey Complex

Figure Test, PPVT: Peabody Picture Vocabulary Test III, Beery VMI: Beery developmental test of visual motor

integration, CST: Card Sorting Test, AF: attention fluctuations

scores, significant impairments were seen in performance on the visual-spatial skills of line

orientation Gudgment of Line Orientation task), visual integration (Rey Complex Figure Test­

copy), and visual motor integration (Beery Developmental Test of Visual - Motor Integration),

on nonverbal long-term memory (Rey Complex Figure Test- delayed recall), sustained attention

(Cancellation Test - speed and attention fluctuations), executive functions (rote memory and

divided attention on Trailmaking Test A and B), and expressive language (Boston Naming Test,

see table 4, all p<0.05). However, children with NF1 scored significantly higher on receptive

language (Peabody Picture Vocabulary Test, z-score=0.31, t=2.21, p=0.031). After correction

for IQ, the visual spatial-skills and nonverbal long term memory were still significantly impaired.

A significant discrepancy between verbal IQ and performance IQ (2:12 points difference) was

noted in 35% of the patients, which was not significantly more in favor of verbal (13%) or

performance IQ (23%) (x2=1.64, p=0.44). Children with a specific comprehensive reading or

mathematics disability had a significant higher frequency of intelligence discrepancies (75%,

x2=6.96, p=0.031, and 70%, x2=6.92, p=0.031).

For 41 children, both the neuropsychological test results and the school performance data on all

four academic fields were present. Children in the No Learning Disabilities group, although with

normal mean IQ (96.8), showed the same dip in their performance IQ profiles (on block design

and object assembly) as the entire NF1 group. Compared to normative scores, children in the

No Learning Disabilities group also scored significantly lower on nonverbal long-term memory

(Rey Complex Figure Test - delayed recall), line orientation Gudgment of Line Orientation

task), rote memory and divided attention (Trailmaking Test A and B), sustained attention

(Cancellation Test- speed and attention fluctuations; all p< 0.05; see table 4).

Disease severity

To test whether there was a relationship between disease severity and cognitive function, we

used the Riccardi Scale, modified to exclude cognitive aspects of NFL A general tendency for

lower scores on didactic and neuropsychological tests was observed with increasing severity of

Cl! \PThR 3 7 3

NF1 (Figure 1). In particular, children with minimal NF1 seemed to consistently perform better

than children with mild or moderate NF1, whose scores were similar on all tests.

A 110

100

~ ~

g~ ~w 80 ~g'

.E 70 "' " ..J

60

50

B 100

80 . 'E: ill !!! 60 c.

" "' .l!! " 40 ~ " 0..

20

0

c 100

95

!!! 90 8 rJl 1: 85 "' "C

" cl3 80

75

70

Technical Reading

#

Special Education

#

Full Scale IQ

Comprehensive Reading

** ###

Learning disabilities

#

VerbaiiQ

Spelling

* ###

Remedial Teaching

Performance IQ

Mathematics

DMinimal NF1 DMild NF1

DMinimal NF1 OMild NF1

• Moderate NF1

0 Minimal NF1 OMildNF1

• Moderate NF1

Figure 1: Relationship between clinical severity of Neurofibromatosis type 1 (modified Riccardi Scale) and learning

efficacy (A), referral to special education, learning disabilities, and remedial teaching (B), and intelligence (C).

Asterisks indicate statistical significant differences using Analysis of Variance or Kruskal-Wallis between the three groups

(* p<O.OS; ** p<O.Ol). Pound signs indicate statistical significant differences using Student's t·test or Chi-Square test

between the minimal NFl group and mild/moderate NF1 groups (# p<O.OS; # # p<O.Ol; # # # p<O.OOS). Bars

represent standard error of the mean. NFl: Neurofibromatosis type 1.

7 4 THE IMPACT OF NEUROFIBROMA.TOSIS TYPE 1 ON SCHOOL PERFOR!Y1ANCE

Indeed, comparison of these two groups revealed that children with mild/moderate NF1 (n=55)

have significantly lower learning efficacies than children with minimal NF1 (n=30) for all 4

didactic domains: technical reading (68 versus 83%, t=2.01, p=0.048), comprehensive reading

(70 versus 95%, t=2.65, p=0.010), spelling (64 versus 82%, t=2.29, p=0.025) and mathematics

(70 versus 94%, t=2.55, p=0.013). In addition, children with mild or moderate NF1 had a

significantly higher frequency of learning disabilities than children with minimal NF1 (86 versus

54%, x2=8.93, p=0.003), received significantly more remedial teaching (94 versus 68%, x2=8.76,

p=0.003), and attended special education more frequently (45 versus 23%, x2=4.05, p=0.044).

The mild or moderate NF1 group had a significantly lower Full Scale IQ (83 versus 92, t=2.19,

p=0.032), verbal IQ (83 versus 93, t=2.33, p=0.023), and Freedom from Distractibility Index

(85 versus 94, t=2.08, p=0.042), as well as significantly lower scores on IQ subtests arithmetic

(p=0.005) and mazes (p=0.028), and test measuring line orientation (p=0.047), divided attention

(p=0.006) and receptive language (p=0.048) than the minimal NF1 group. There was no

significant difference between familial or sporadic NF1 on any parameter of school

performance.

Discussion

To our knowledge, this is the first study in which the impact of NF1 on school performance is

determined by combining quantitative data on didactic performance obtained in the school

setting with information on special education and remedial teaching. Our results clearly

demonstrate that school performance is severely affected in children with NFL At least 75% of

the children with NF1 have one or more learning disabilities in technical reading,

comprehensive reading, spelling or mathematics. The learning disabilities are distributed equally

over these 4 didactic domains, indicating that NF1 pathology does not cause one specific type of

didactic deficit. The high incidence of remedial teaching (85%), special education (37%) and

grade repetitions (40%) in our study emphasizes the school problems arising from NF1. Only

10% of the children do not have problems in any aspect of school functioning. School

performance and cognitive functioning were found to be substantially more affected in patients

with more severe physical features ofNF1.

The 75% learning disabilities found in this study group is markedly higher than reported

previously (52% learning disabilities)6, but are in agreement with the 70% total learning

disabilities of Brewer et aLB The frequency of learning disabilities found in our population

reflects actual problems experienced by children with NF1 in their school career. This high

C~L\P'lT:R 3 7 5

number is likely to be more representative because the didactic tests obtained from the school

setting that were used in our study have a higher ecological validity than academic achievement

tests used in other studies. The Dutch Student Monitoring System assesses didactic progress of

each student 3 times a year throughout the school career and scores are compared to normative

grade peer scores. Although a performance of -1 SD as a cutoff for learning disabilities was also

used in the studies mentioned above, it is still an arbitrary definition. However, even when using

a performance of -2 SD as a cutoff ~earning efficacy < 70%), still 63% of the children with

NFl display impairments in one or more didactic domains. Notably, the incidence of 75%

learning disabilities could be an underestimation of the real school problems experienced by

these children, because we have shown that although some children do not display learning

disabilities in their didactic scores, they obtain these scores only in the context of remedial

teaching specifically for problems in learning. In addition, children in the No Learning

Disabilities group are still significantly impaired in nonverbal long-term memory, executive

functions and attention. Mild and specific learning disabilities tend to become apparent when

increasing demands on cognitive function can no longer be met. This phenomenon resembles

'growing into deficit'.21 Thus, children without learning disabilities can grow into specific

learning disabilities at an older age. This is supported by the analysis that children without

learning disabilities in our study are significantly younger than children with learning disabilities

(2.4 years, p<O.OOOS). Our data further indicate that the impact of NFl on school is not only

limited to cognitive function, but encompasses motor function as well. In total, 52% of the

children received remedial teaching for physical problems such as motor problems or speech

problems.

Disease Severity

The phenotype of NF1 is highly variable, even in patients with identical NFl mutations, which

is proposed to be due to genetic modifiers.22 The relationship that we found between physical

symptoms and cognitive symptoms suggests a common genetic basis for both. So far, this

relationship has only been observed for children with seizures,23 and for patients with a

microdeletion.24 However, the more severe cognitive phenotype of this latter group may be

caused by the deletion of genes outside the NFl gene. The lack of a correlation between disease

severity and cognitive function in other studies could be caused by not applying a severity

scale,22 by excluding patients with specific NFl characteristics (for instance, patients with optic

gliomas 25) or by using global intelligence instead of more detailed cognitive functions.zs

7 6 THE IMPACT OF NEUROFIBRO!vL-\TOSIS TYPE 1 ON SCHOOL PERFOR!vL~NCE

Neuropsychological testing

We found a markedly higher frequency of mental retardation in our sample than previously

reported in other studies on NF1 (4-8%)7. However, if we shift the normal distribution curve of

IQ to the left to an average IQ of 86 (with a standard deviation of 15) as found in our study, the

expected frequency of mental retardation would be 14.3%, which is in accordance with our

findings. Interestingly, in our study group only 3 children (4.8%) had both a performance IQ

and a verbal IQ of below 70, indicating that not all of the children with a full scale IQ below 70

perform at the level of mental retardation.

Based on the neuropsychological profile of NF1 patients, remedial teaching for learning

disabilities in children with NF1 could be tailored to address the specific cognitive functions that

are impaired. In addition, weaknesses that are specific to NF1 can be avoided. For instance,

many children with NF1 have a poor visual analysis and could benefit from a verbal rather than

visual presentation of mathematical problems. The potential value of remedial teaching is

illustrated by the observation that most children in the No Learning Disabilities group receive

remedial teaching. Thus, despite the observed deficits in the neuropsychological profile of these

children, they do not (yet) show deficits in any of the didactic domains. However, as discussed

above, there is a fair chance that these deficits will become apparent at older age. Thus, although

remedial teaching and special education can ameliorate learning challenges, these interventions

are not sufficient to eliminate them.

Our study shows large attention problems in children with NFL Hyman et al.6 showed

comorbidity of attention-deficit hyperactivity disorder with literacy problems. The importance

of attention for technical reading is supported by recent evidence that methylphenidate

improves comorbid dyslexia in children with attention-deficit hyperactivity disorder.26

Methylphenidate has been reported to have favorable effects on attention and behavior in NF1

patients.27 As noted above, the children in the No Learning Disabilities group show severe

attention deficits, which could put them at risk for developing learning disabilities over time.

Although these children could potentially profit from timely recognition and treatment with

methylphenidate, only one child in this group received medication for attention-deficit

hyperactivity disorder, which illustrates the risk to overlook attention problems in children with

relatively good school results.

(R\PTER3 7 7

Limitations

The patients participating in this study were selected from the patient group of a university

hospital with a specialized NF1 clinic. Potentially, this could result in a referral bias toward

children with more severe physical symptoms. However, the large group of children with only

minimal NF1 contradicts a referral bias. Also, there was no difference between the group for

which a school performance questionnaire was received (n=86) and the group for which it was

not (n=28) regarding the distribution of disease severity (p=0.28), age (p=0.86), the frequency

of special education (p=0.63) or the frequency of mental retardation (p=0.68). This indicates

there was no selection bias in the data from the school performance questionnaire. Finally, the

frequency of mental retardation does not differ between the group that consented to

neuropsychological assessment (n=62) and the group that did not (n=52) (p=0.14; data from

the non-response group was obtained from patient charts).

It should be noted that many school performance questionnaires missed quantitative data on

one or more of the four didactic domains. In most of the cases that data were missing, teachers

did provide qualitative data; however this was not used in our study. Evaluation of these

qualitative scores does not suggest a bias toward selectively omitting good or bad school

performance data. However, missing data could lead to an underestimation of the amount of

learning disabilities, because children who show no learning disabilities but have missing data for

some domains (n=7), could still have learning disabilities in the missing domains. Therefore, we

have stated that the percentage of learning disabilities we found is at least 7 5%. In the analysis of

disease severity, we did not exclude children with a microdeletion because not all children

received genetic testing. This could potentially influence the strength of the association found in

our study. However, the severity scores of the 4 children in our study that had a known

microdeletion were evenly distributed over the severity groups (2 minimal, 1 mild and 1

moderate NF1), making a confounding effect less plausible.

The high number of children with NF1 who receive special education or remedial teaching

found in our study population is alarming but is also encouraging because it suggests that the

Dutch school system does recognize the need for extra support of children with NF1. Because

school systems may be organized differently in other countries, the actual percentage of children

receiving special care may vary from country to country. However, the reported odds ratio for

receiving special education or remedial teaching should be applicable to all school systems.

Therefore, our study can be used as a general guide to counsel parents and teachers to be alert

7 8 THE IMPACT OF NElTROFIBROMATOSJS TH'E 1 ON SCHOOL PERFORMANCE

for problems in learning, motor functioning, speech, and behavior. Awareness of the school

problems associated with NF1 may facilitate timely recognition and appropriate support.

Acknowledgments

We are very grateful to all the children with NF1 and their parents and teachers for their

participation and to E. Barendse, M.J. Bouman, A. Lauxterman-Gaemers, and R. Wierenga for

their contribution to the neuropsychological assessment. We greatly appreciate the support of all

participants of the NF1 clinical work group from the departments of Pediatric Dermatology,

Pediatric Ophthalmology and Clinical Genetics, and all members of the NF1 CoRe (Cognitive

Research) team.

Cr!\PT!R3 79

References

1. Huson SM, Hru:per PS, Compston DA. Von Recklinghauscn neurofibromatosis. A clinical and population study in

south-east Wales. Brain 1988;111 ( Pt 6):1355-1381.

2. Cnossen MH, Moons KG, Garssen MP, et al Minor disease features in neurofibromatosis type 1 (NF1) and their

possible value in diagnosis of NF1 in children < or = 6 years and clinically suspected of having NFL

Neurofibromatosis team of Sophia Children's Hospital. J Med Genet 1998;35:624-627.

3. North K, Hyman S, Barton B. Cognitive deficits in neurofibromatosis 1.] Child Neurol 2002;17:605-612; discussion

627-609, 646-651.

4. Ozonoff S. Cognitive impairment in neurofibromatosis type 1. Am J Med Genet 1999;89:45-52.

5. Barton B, North K Social skills of children with neurofibromatosis type 1. Dev Med Child Neurol 2004;46:553-563.

6. Hyman SL, Arthur Shores E, North KN. Learning disabilities in children with neurofibromatosis type 1: subtypes,

cognitive profile, and attention-deficit-hyperactivity disorder. Dev Med Child Neurol2006;48:973-977.

7. North KN, Riccardi V, Samango-Sprouse C, et al. Cognitive function and academic performance in

neurofibromatosis. 1: consensus statement from the NF1 Cognitive Disorders Task Force. Neurology 1997;48:1121-

1127.

8. Brewer VR, Moore BD, 3rd, Hiscock M. Learning disability subtypes in children with neurofibromatosis. J Learn

Disabil 1997;30:521-533.

9. National Institutes of Health Consensus Development Conference Statement: neurofibromatosis. Bethesda, Md.,

USA, July 13-15, 1987. Neurofibromatosis 1988;1:172-178.

10. Riccardi VM. Nettrofibromatosis: Phenotype, Natural History, and Pathogenesis. 2nd ed. Baltimore: Johns Hopkins University

Press; 1992.

11. Achenbach TM. Manual for Teacher's Report FomJ & 1991 prufi!e. Burlington: University of Vermont; 1991.

12. Melis G. DIE Boek. Lisse- Leeuwarden: Eduforce - Swets & Zeitlinger B.V.; 2003.

13. Arts M. Indicatieste!ling en m'teria voor het spedaal onderwijs of een 171gzak, brochure voor ouders. Zoetenneer: Speed Print; 2005.

14. Dutch Ministry of Education, Culture and Science. Wat basisscholen moeten weten over indicatiestellingpraktijkonderwijs en

leerwegonderstemtend ondenvijs in het sehoo!faar 2002-2003. Den Haag: Dutch Ministry of Education, Culture and Science;

2002.

15. Lezak MD, Howieson DB, Loring DW, et al. Neuropsychological Assessment. 4th ed. New York: Oxford University

Press; 2004.

16. DeBruyn EEJ, Vander Steene G, Van Haasen PP. Wechsler Intelligence Scale for Children- Revised, Dtttch edition. Lisse:

Swets & Zeitlinger; 1986.

17. Dutch Ministry of Education, Culture and Science. Regeling lijst van te gebruiken instrumenten bij indicatiestelling

voor leerwegondersteunend onderwijs (LWOO) en praktijkonderwijs (PRO) voor instroom schooljaar 2005-2006.

Ge!e Katern 2004;13:26-31.

18. Dutch Ministry of Education, Culture and Science. Feiten en Cijfers 2006 Web site. Available at:

http:/ /www.cijfers.rninocw.nl. Accessed May 29, 2007.

19. Dutch Inspection Of Education. De staat van het onderwijs, ondenvijsverslag 2004-2005. Den Haag: Den Haag Media

Groep; 2006.

20. Mathijssen J, Jeeninga, W. Brabantse ]ettgdmonitor enqui'te 0- t/ m 11 jarigen: Tabellenboek Provinde Noord-Brabant. Den

Bosch: GGD Hart voor Brabant; 2006.

21. Aarsen FK, Paquier PF, Redtlingius RE, et al. Functional outcome after low-grade astrocytoma treatment in

childhood. Cancer 2006;1 06:396-402.

8 0 THE IMPACT OF NEUROFIBROI.\iATOSIS TYPE 1 ON SCHOOL PERFORI.\LI>NCE

22. Easton DF, Ponder JVIA, Huson SM, et a!. An analysis of variation in expression of neurofibromatosis (NF) type 1

(NF1): evidence for modifying genes. Am J Hum Genet 1993;53:305-313.

23. Szudek J, Birch P, Riccardi VM, eta!. Associations of clinical features in neurofibromatosis 1 (NF1). Genet Epidemio!

2000;19:429-439.

24. Descheemaeker MJ, Roelandts K, De Raedt T, et a!. Intelligence in individuals with a neurofibromatosis type 1

microdeletion. Am J Med Genet A 2004;131 :325-326.

25. Hyman SL, Shores A, North KN. The nature and frequency of cognitive deficits in childten with neurofibromatosis

type 1. Neurology 2005;65:1037-1044.

26. Keulers EH, Hendtiksen JG, Peron FJ, et a!. Methylphenidate improves reading performance in childten with

attention deficit hyperactivity disorder and comorbid dyslexia: An unblinded clinical trial. Eur J Paediatr Neuro!

2007;11:21-28.

27. Mautner VF, Kluwe L, Thakker SD, eta!. Treatment of ADHD in neurofibromatosis type 1. Dev Med Child Neuro!

2002;44:164-170.

CH.\PTER 3 8 1

, Lianne C. Xtab, MScl.3, R. Oostenbrink, MD PhD1,Atja de Goede-Bolder, MD1,

Femke K. Aa:rsen, MA2, Ype Elgersma, PhD3, Heru:iette A. Moll, MD lhD1

"' ~' "'

The NFJ CoRe Team (Cognitive Resc:trch Tc:un}, Erasmus A-'IC/Sophia Cbildrcn's Hospital, Rotterdam, Tbe Netberlands. 1Department of General Pediatrics,

2 Department of Pediatric Neurology, 3 Department of Neuroscience.

Journal of pediatrics, in press (2008)

Abstract

Objective

We aim to investigate Health Related QOL (HR-QOL) in children with Neurofibromatosis type

1 (NF1) using parental reports and children's self-reports, and to investigate the potential

contribution of demographic factors, disease-specific factors, and problems in school

performance or behavior.

Study Design

In a prospective observational study, parents of 58 children with NF1 (32 boys, 26 girls, age 12.2

± 2.5 years) visiting a university clinic, and their 43 children 10 years or older were assessed with

the Child Health Questionnaire (CHQ). Potential determinants of domain scores were assessed

in three explorative regression models.

Results

Parents reported a significant impact of NF1 on 9/13 CHQ scales, with moderate effect sizes

on 8 (General Health Perceptions, Physical Functioning, General Behavior, Mental Health, Self

Esteem, Family Activities, Role functioning Emotional/Behavioral, and Parent Emotional

Impact). Children report an impact on Bodily Pain, and an above average General Behavior.

Multiple CHQ scales were sensitive to demographic factors and behavioral problems, and one

to NF1 severity. NF1 visibility and school problems did not influence HR-QOL.

Conclusions

Parents, but not NF1 children themselves, report a profound impact ofNF1 on physical, social,

behavioral and emotional aspects of HR-QOL. Multiple HR-QOL domains were most sensitive

to behavioral problems, which points to an exciting potential opportunity to improve HR-QOL

in children with NF1 by addressing these behavioral problems .

8 4 HEALTH RELATED QUALITY OF LIFE IN CHILDERN WITH NEUROFIBRO:MATOSIS TYPE 1

Introduction

Neurofibromatosis type 1 (NF1) is an autosomal dominant genetic disease with an incidence of

1 in 3000, half of which are de novo cases.1· 2 The disease is characterized by various progressive

neurocutaneous manifestations, including cafe au lait maculae, axillary freckling, neurofibromas,

and Lisch nodules. At pediatric age, possible complications include deformities due to plexiform

neurofibromas, neurologic problems, malignancies, endocrine disturbances and orthopedic

problems such as scoliosis.z However, the most frequent complication of NF1 in children is

cognitive impairment, characterized by a low-average IQ, and problems with visual-spatial skills,

memory, language, executive functioning and attention). 4 Due to these problems, up to 7 5% of

the NF1 children have learning disabilities and the majority of children needs additional support

in the form of special education or remedial teaching.4 In addition, children with NF1

demonstrate poor social skills, are frequently picked on by peers and have fewer friends.s-7

Behavioral problems are commonly reported, and include Attention Deficit Hyperactivity

Disorder (ADHD) in up to 40%.5. 6, s. 9

As can be expected from the various physical, cognitive and social complications associated with

NF1, a below average Quality of Life (QOL) has been reported by NF1 adults.IO, n This lower

reported Quality of Life was recently confirmed for NF1 children in one study on preschoolers,

and one on children aged 7 to 16 including child self-reports.12, 13 Although NF1 physical disease

severity and disease visibility have been related to lower scores on QOLJO, 12, 13, it has not been

investigated to what extent the school problems and behavioral problems experienced by

children with NF1 contribute to problems reported in QOL.

This study aims to investigate Health Related QOL (HR-QOL) in children and adolescents with

NF1 using parental reports and children's self-reports, and to investigate the potential

contribution of demographic factors, disease-specific factors, and problems in behavior or

school performance.

Methods

Procedure

We performed this prospective observational study in the context of the baseline inventory of a

larger study on cognitive functioning among 62 children with NF1 between January 2006 and

Cr-r\Pnn~+ 8 5

March 2007.14 Participants were recruited from the multidisciplinary pediatric NFl-outpatient

clinic of the Erasmus MC - Sophia Children's Hospital Rotterdam, which is a supraregional

reference center covering about 3 million citizens. For detailed inclusion and exclusion criteria

we refer to our previous publication.4 For the purpose of the current study, patients were only

included if at least the parent HR-QOL questionnaire was completed. Children visited the clinic

for neuropsychological testi.ng.4 During this visit, children and parents received questionnaires

on HR-QOL, and an additional questionnaire on behavior, to be completed by the child's

teacher. A return envelope was provided to allow for filling out the questionnaire at home. This

study was approved by the Erasmus MC - Sophia Children's Hospital Medical Ethical

Committee.

Measurements

Health related Quality of Life (HR-QOL)

HR-QOL was assessed with the Child Health Questionnaire (CHQ).15 This internationally

applied generic questionnaire covers physical, psychological and social aspects of quality of life.

It comprises a Parent Form (CHQ-PFSO; 50 items, 13 domains), to be completed by parents of

children from 5 to 18 years old and a Child self-report form (CHQ-CF87; 87 items, 12 domains)

to be completed by children from 10 years old themselves. Reference-values and reliability and

validity of the CHQ-PFSO and CHQ-CF87 have been determined for the Dutch population_l6, 17

Parallel domains in the CHQ-PFSO and CHQ-CF87 are 8 multi-item scales: General Health

Perceptions, Physical Functioning, Bodily Pain, General Behavior, Mental Health, Self Esteem,

Family Activities, and Role Functioning - Physical, and 2 single item questions: Family

Cohesion and Change in Health. The child form further provides a Role Functioning - Behavior

and Role Functioning - Emotion score, and the parent form a Role Functioning -

Emotion/Behavior summary score, Parental Time Impact and Parental Emotional Impact scale.

Role Functioning refers to limitations in schoolwork or activities as a result of behavioral

problems, emotional problems or both. Scale item scores are summed and transformed into

scores on a scale of 0 (worst possible health state) to 100 (best possible health state).IS

Demographic and disease-related factors

All clinical data were registered by an experienced pediatrician of the NFl team (A. de G-B).

Familial or sporadic NF1 was recorded. Socio-economic status (SES) was determined from

highest parental occupation or, if not present, highest parental education, and divided into low,

middle or high (modified from a standard occupation classification) .IS Based on the last visit to

the outpatient clinic, visibility of NF1 when fully dressed was scored according to Ablon,19 with

8 6 HEALTH RELATED QUALITY OF LJFE IN CHJLDERN WITH NEUROFIBROMATOSIS TYPE 1

mild visibility indicating no visible tumors and unremarkable gait and posture, moderate

visibility indicating patients had some visible tumors or skeletal features without noticeable limp,

and severe visibility indicating the presence of numerous visible tumors, optic glioma affecting

sight, or severe skeletal features with noticeable limp. Physical disease severity was scored

according to the most recent version of the Riccardi Scale,20 modified to exclude cognitive

aspects of NF1 in order to be able to assess physical severity and cognitive problems separately.

Severity was scored Minimal (when the patient had no features that compromise health, i.e. only

harmless cosmetic features such as cafl au fait maculae, freckling and Lisch nodules), Mild

(patient had minor medical complications such as ptosis or discrete plexiform neurofibroma),

Moderate (patient had complications that are a significant compromise to health, such as

paravertebral neurofibromas or low grade glioma) or Severe (medical history of malignancy).

Behavior and School performance

We used the Total Problems score on the standardized Teacher's Report Form (TRF, 118

items)21 to assess behavioral and emotional problems. The TRF is validated for the Dutch

population and provides summary scores for Internalizing (subscales social withdrawal, somatic

complaints, and anxiety/ depression), Externalizing (subscales rule breaking behavior and

aggressive behavior), and Total problems (overall summary score). Items are rated 0 (never

true), 1 (sometimes true) or 2 (clearly or often true). Scores are converted toT-scores (mean 50,

SD 10), with higher scores indicating more problems.

We used data on school type, learning disabilities and need for remedial teaching, obtained from

teacher questionnaires in the ongoing study on cognition,4 to obtain a 4-level school-scale: 1)

No learning disabilities, no remedial teaching, regular education; 2) Learning disabilities and/or

remedial teaching, regular education; 3) School for the learning disabled; 4) School for the

mentally retarded or severely learning disabled.

Analysis

Data were analyzed in SPSS 12.0. Differences in CHQ scores compared to reference values16, 22

were assessed using a two-sided independent t-test. For each CHQ domain, effect sizes were

calculated compared to reference values16, 22 in order to evaluate clinical relevance of scores

(mean reference - mean NF1)/ (square root (pooled SD)). According to Cohen's guidelines,

effect sizes from 0.2 to 0.5 were defined as small, from 0.5 to 0.8 as moderate, and >0.8 as

large.23

CH.\PTbR+ 8 7

Internal reliability of the CHQ domains when applied to the NF1 population was assessed using

Cronbach Alpha. To compare children's and parent's ratings, we looked at differences in effect

sizes, since absolute differences in scale scores are not only informative of NF1 but also reflect

differences observed between parents and children's ratings in the normal population. In

addition, to examine the strength of concordance between ratings by parents and children we

calculated Intraclass Correlation Coefficients (ICC's), which take into account the individual

variability between parent and children pairs. ICC's below 0.4 were considered to reflect poor to

fair, between 0.4 and 0.6 moderate, between 0.6 and 0.8 good and above 0.8 excellent

agreement.24 Because the CHQ-CF is constructed for children aged 10 years or older,

comparison of children's and parent's ratings is performed on the subgroup of children aged 10

years and older only (n=43).

Using multiple linear regression (enter method) we built three separate models to explore

determinants of CHQ domains scores that were significantly affected. Model 1 contained

demographic factors (child's age, child's sex, socio economic status of the family,

familial/sporadic NF1). Model 2 included disease related factors (NF1 severity and visibility).

Model 3 consisted of problems in behavior and school performance (TRF total problems,

School Scale). For reliable analysis, scales with less than 5 patients on a subscale (visibility

severe, severity severe, School Scale 1) were merged with the closest ranking subscale. The

patient with unconftrmed familial NF1 was excluded from analysis of model 2. Reported values

are the regression coefficients for the condition, corrected for the other conditions within the

model.

Results

The CHQ-PF was received from parents of 58 out of 62 eligible children (response rate 94%).

Of these 58 children, 43 were aged 10 years or older. All of these 43 children completed the

CHQ-CF (response 100%). In addition, the TRF was received from 54 of the teachers of the 58

participating children (response 93%). The study included four sibling pairs. Patient

characteristics are shown in table 1. Complications resulting in a Mild severity score were

constipation (n=4), discrete plexiform neurofibroma (n=10), sleep disturbance (n=2), ptosis

(n=2), strabismus (n=1), scoliosis (n=1), leg length asymmetry (n=1), deafness (n=1), or a

Central Nervous System cyst (n=1 ). Moderate severity was scored for paraspinal neurofibromas

(n=9), diffuse plexiform neurofibroma (n=3), puberty disturbance (n=5), low-grade glioma

8 8 HEALTH RELATED QUALITY OF LIFE IN CH!LDERN WITH NEUROFIBROMATOSIS TYPE 1

(n=3) or pseudo arthrosis (n=l). One child had a Severe severity due to a medical history of

Myelo-Dysplastic Syndrome.

Table 1: Characteristics of children with NF1 (N~SS) and their parents.

Characteristic

Demographic factors

Age child (mean)

Male sex

Socio-Economical Status

Low

Middle

High

Mode of inheritance

Familial

Sporadic

unconfirmed

Comorbid conditions

Disease-specific factors

Ablon Visibility Scale

Mild

Moderate

Severe

Modified Riccardi Scale

Minimal

Mild

Moderate

Severe

School performance and behavior

School Scoret

1-No LD,no RT

2 - LD and/ or RT, regular education

3 - School for the learning disabled

4 -School for the mentally retarded

Behavior (TRF Total Problems, T -score)

*p<0.05 compared to normative scores (mean- 50, SD - 10).

N(SD)

12.2 (2.5)

32

21

17

20

22

35

7

44

10

4

13

23

21

3

27

18

7

56.2 (8.6)*

%

55

36

29

35

38

60

2

12

76

17

7

22

40

36

2

6

49

33

13

tfor 3 patients on regular education, information on learning disabilities or remedial teaching was missing.

NF1, Neurofibromatosis type 1; LD, Learning Disabilities; RT, Remedial Teaching; TRF, Teacher's Repon Form.

Reported CHQ domain scores are shown in table 2. Parents rate their children's HR-QOL as

significantly lower than reference values for 9 out of 13 domains, of which 8 with a moderate

effect size (General Health perceptions, Physical Functioning, General Behavior, Mental Health,

Self Esteem, Family Activities, Role Functioning Emotional/Behavioral, and Parent Emotional

Cl-f\PmR-1 8 9

Table 2. Parent- and child reported CHQ scores for children with NFL

Parental Report (N-58) Child Report (N=43)

NFI Referencet NFI Reference*

Scale (range 0-100) Mean (SD) l\Iean (SD) Effect size Mean (SD) Mean (SD) Effect size

General Health Perceptions 71.9** (17.5) 82.9 (13.4) -0.7 72.7 (16.2) 74.6 (15.9) -0.1

Physical Functioning 94.5** (9.5) 99.1 (4.3) -0.6 95.7 (8.9) 96.8 (5.4) -0.1

Bodily Pain§ 81.2 (17.4) 85.7 (17.2) -0.3 71.4* (27.5) 78.2 (19.5) -0.3

General Behavior 69.5** (17.6) 78.5 (13.1) -0.6 86.8* (8.8) 83.6 (10.2) 0.3

Mental Health 75.3** (14.7) 81.4 (12.1) -0.5 79.8 (12.8) 78.2 (13.0) 0.1

Self Esteem 72.9** (12.8) 79.2 (11.0) -0.5 76.9 (13.5) 75.4 (12.5) 0.1

Family Activities 77.8** (23.1) 91.5 (11.9) -0.7 83.0 (17.5) nr nr nr

Family Cohesion 67.2 (23.8) 72.2 (19.4) -0.2 75.7 (26.7) 75.7 (23.1) 0.0

Change in Health§ 59.5 (16.8) nr nr nr 66.1 (22.6) nr nr nr

Role Functioning- Physical 94.0 (15.2) 95.8 (15.6) -0.1 96.1 (11.3) 96.5 (11.6) 0.0

Role Functioning- Emotional na na na na na 92.0 (14.4) 92.3 (16.8) 0.0

Role Functioning- Behavioral na na na na na 93.0 (17.1) 91.4 (13.7) 0.1

Role Functioning- Emotional / Behavioral 90.4** (17.0) 97.9 (7.2) -0.6 na na na na na

Parental Emotional Impact 73.0** (21.1) 86.3 (15.2) -0.7 na na na na na

Parental Time Impact 87.5** (18.2) 94.0 (13.0) -0.4 na na na na na

*=p<0.05, ** p<0.01 for comparison NF1 scale scores to reference scale scores.

tReference population: parents of Dutch schoolchildren 5-13 years, N=353.'5

+Reference population: Dutch schoolchildren 9-17 years, n=444.''

§=Child report n=42.

nr: No reference values available, na: Not applicable.

Impact). Parental reports for children older than 10 years were similar to parental reports for

younger children. Children rated their HR-QOL not different from reference values, except for

significant lower scores on Bodily Pain and significant higher scores on General Behavior, both

with a small effect size. For thls NF1 population, internal reliability of the CHQ parent and

children scales was very good (Cronbach's Alpha 0.73-0.93) except for General Health

Perceptions, Physical Functioning and Self Esteem reported by parents (0.54 to 0.66).

As shown in figure 1, we observed moderate to large differences between parental and

children's ratings on the majority of the domains (differences more than 0.5 effect size in 5 out

of 8 scales depicted), with generally higher scores, indicating less problems, reported by children

than by their parents. ICC's between parent and children's ratings ranged from poor (2 scales) to

moderate (4 scales) and good (2 scales). The size of the gap between effect sizes reported by

parents and children did not always match the level of concordance. On Physical Functioning,

the difference between parents and children was moderate (0.5 effect size), but there was a good

concordance (0.72). This should be interpreted as follows: both parental and child reported

CHQ scores deviate from reference values in the same direction (i.e. both lower scores) but

nevertheless there is a moderate absolute difference in the domain scores reported by parental

and children. For General Behavior, the difference between parental and children reports is

large (0.9 effect size) and the concordance is poor (0.30), indicating a high variability between

parental and children's reports. Family activities (ICC 0.59) and Change in Health (ICC 0.36) are

not depicted in figure 1 because no reference values were available.

To explore which determinants are related to HR-QOL scores, we performed a regression

analysis in three models: demographic factors, disease-related factors, and problems in school

performance or behavior (table 3). Exploration of demographic factors in model 1 revealed that

high SES contributes negatively to the score for Bodily Pain in children. For parents, we found a

significant positive impact of male sex on Parent Time Impact, and a positive influence of

familial NF1 (on Self Esteem). Age of the child did not influence CHQ domain scores. In

model 2 (disease-related factors), we observed a negative relationship between severity and

General Health Perceptions (significant for moderate versus minimal severity), but no

significant influence of NF1 visibility. In model 3, multiple CHQ scales were sensitive to

behavioral problems reported by teachers. Total problems on the TRF had a negative impact on

General Health, General Behavior and Parent Time Impact. School performance did not

influence any CHQ domain score.

CH\PTER4 9 1

0,8 0,7

Ill 0,6 8 0,5

~ 0,4

5 0,3 c 0,2

111!! 0,1 IIIQI

Comparison of CHQ scores of parents and children

-a- Children ....... Parents -Reference

9.'e 0,0 IE 0 -0,1 w; -0,2

__ --!'!'~ ical Bod II P'!ih General Ger era I --_ / behavior he ~th funct oning ----

Mental Self e steam Fa ily coh !'i9D~~~~R le-

I!! -0,3 :!. -0,4

~ -0,5 u -0,6

Hell th ph sical ',,',, 1 ~~ --- f //// .......... ,.......... ---- /

~~

-0,7 -· -0,8 *** -0,9

-1,0 *** ICC: 0.30 0.72 0.51 0.30 0.53 0.46 0.38

Figure 1: Comparison of CHQ scores of NF1 parents and children.

Error bars represent 95 percent confidence intervals. Differences compared to table 2 are due to rounding.

CHQ: Child Health Questionnaire, NFl: Neurofibromatosis type I, ICC: Intraclass Correlation.

Discussion

0.69

We studied the impact of NF1 on HR-QOL in a large group of children with NF1, using

parental ratings complemented with children's self reports. Parents report that children with

NF1 experience substantial problems compared to healthy children on 9 out of 13 CHQ

domains, with moderate effect sizes on 8 domains. We observed profound impact of NF1 on

parents themselves and on their family life. In contrast, we observed that children with NF1

report lower scores on Bodily Pain only, and report above average scores on General Behavior.

We measured substantial differences between the effect sizes of parental and children's ratings

on the majority of cross-compared scales. Although parents usually reported larger impairments

than children, their scores tend to correlate. Social-economic status, sex, familial NF1, NF1

severity and in particular the presence of behavioral problems influenced several CHQ domains.

None of the evaluated determinants influenced child reports for General Behavior, or parental

reports for Physical Functioning, Role Functioning Emotional/Behavioral, Mental Health or

Family activities.

The substantial impairments over multiple HR-QOL domains reported by parents of NF1

children in our study confirm results of other studies using parental reports of QOL in children

with NF112. 13, and self reports of NF1 adults.1D. 13 In our study, children reported lower scores

9 2 HEALTH RELATED QUALITY OF LIFE IN CHILDERN WITH NEUROFIBROJ\-L"'-TOSIS TYPE 1

Table 3: summary of outcome explorative regression analysis on Child Health Questionnaire (CHQ) scales

MODELl

Age child

Gender

SES (high)

Familial NF1

MoDEL2

Severity (mod/severe)

Visibility

MODEL3

Behavior (fRF) t

School Scale

CHILD SELF REPORT

BP GB

-26.6±10.2'

PARENTAL REPORT

GH PF GB PT

10.8±4.7'

8.9±5.1§

-14.1±6.4'

-0.6±0.3' -0.7±0.3' -0.7±0.3'

PE REB IviH SE FA

11.2±6.01 7.3±3.6'

Values (regression coefficients ± SD) represent difference in CHQ scale score in points (1-100), per unit of increase in score for the independent variable (continuous variables), or compared to

the reference category of the variable (female sex, low SES, minimal severity, school scale category I/II, spontaneous NF I, mild visibility).

+:positive impact,-: negative impact on CHQ score, empty box: no association.

*p<0.05, §p<O.l.

tHigher scores on the TRF indicate more behavioral problems.

Mod.: moderate, BP: Bodily Pain, GB: General behavior, GH: General health perceptions, PF: Physical functioning, PT: Parental time impact, PE: Parental emotional impact, REB: Role

functioning- emotional/ behavioral, l'viH: l\{ental Health, SE: Self esteem, FA: Family activities, SES: Socio-economical Status, NFI: Neurofibromatosis type 1, TRF: Teacher's Report Form.

on one HR-QOL domain only, which is in contrast to the only other study that has investigated

children's self ratings of Q0Ll3 , showing problems in motor, cognitive, social and emotional

domains of the TNO-AZL Child Quality of Life Questionnaire (TACQOL). This discrepancy

with our results may be explained by methodological differences between the CHQ and

TACQOL. CHQ scores are based on the reported frequency of problems 15 , whereas

TACQOL domain scores are based on the emotional distress due to the problem rather than

the frequency.25 Possibly, children with NF1 less frequendy report or perceive encountered

problems than parents, but do report emotional distress over these problems. In our study, we

preferred the use of the CHQ because it incorporates 4 domains to assess the impact of a

disease on the parents themselves and on the family as a whole, which is not covered by the

TACQOL.

The striking above average self-ratings of General Behavior by children are refuted by the

substantial impairments in behavior reported by teachers on the TRF, but also by objective

measurements of attention reported for this patient group in our previous study.4 This over­

estimation is in line with reports of above average self-concept in NF1 adults,26 above average

self-perceived academic achievement27 and social skills 7 in NF1 children, and discrepancies

between child and parent perceived NF1 disease severity.2B Together, these reports strongly

suggest that children with NF1 have problems in forming or reporting an accurate self-concept.

Over-estimation in deficient area's could be related to factors observed more generally in

children with learning disabilities or ADHD, such as self-protective mechanismsp. 29, 30 or to

poor cognitive skills in general_31 Alternatively, NF1-specific cognitive impairments might also

include impairments in self percept an sich. In addition to over-estimation of General Behavior,

the paucity of problems reported by NF1 children sharply contrasts to the substantial problems

reported by parents on multiple CHQ domains. Large differences between parental and child

ratings, with higher ratings in children, have been widely reported in other chronic conditions,

such as ADHD and cerebral palsy_32, 33 Interestingly, in ADHD children, discrepancies between

parental and child self reports increased with ADHD symptoms of inattentive and combined

ADHD.s2 Since the vast majority of ADHD in NF1 is of these subtypess, this may pardy

explain differences between parental and children's ratings in our study.

We observed sensitivity of CHQ domain scores to several of the potential determinants. The

negative impact of SES on CHQ domain scores may be explained by higher expectations for

their child of parents with higher SES, and the experienced discrepancy between parental

expectations and the actual level of functioning of their child. Contradictory results however,

9 4 HEALTH RELATED QlJALTTY OF LIFE IN CI-IILDERN WITH NEUROFIBROMATOSIS TYPE 1

were observed in younger NF1 children, where educational level (not occupation) of the

respondent contributed to higher HR-QOL on 5 out of 11 domains, as well as in general

population studies.12, 34 . The positive impact of familial NF1 on CHQ scores is also observed

in the study on toddlers_l2 Experience with NF1 in the family may lead to better coping styles,

or less recognition of problems in these domains. We could not reproduce the reports of

previous studies on NF1 children of an impact of severity and visibility on the emotional

domains,B of parent-rated severity on Parent Emotional Impact, 12 and visibility on General

Health Peceptions.12 However, in the latter study,12 both disease severity and HR-QOL were

scored by the parents themselves. Parental perspectives influencing both severity score and HR­

QOL score could seriously confound the measured relationship. Substantial impact of both

severity and visibility on QOL is reported by NF1 adults using self-scored!! and physician

scored10 severity and visibility.10. 11 NF1 is a progressive disease, and many children do not yet

display the cutaneous signs adults do. Therefore, it may be more informative to assess the

impact of visibility and severity of NF1 on QOL in adults.

Our study is the first to address behavioral problems and school problems as potential

determinants of HR-QOL. Our explorative regression analysis shows that problems in the

behavioral domain are important determinants of HR-QOL scores. The negative impact of

behavioral problems seems plausible considering the demand put on parents of children with

behavioral disturbances. The sensitivity of multiple HR-QOL domains to behavioral problems

underlines the importance of adequately managing the frequent behavioral problems of children

with NF1, for instance with stimulant medication,35 social training programs, and better

education of families. So far, studies have only focused on fixed factors influencing HR-QOL

such as familial NF1, sex, SES and disease severity. Thus, our study offers a first potential

handhold for improving HR-QOL. The effect of behavioral therapies on HR-QOL in NF1

should be addressed in future studies.

The lack of an impact of school performance on any CHQ domain may indicate that parents

and children do not perceive school problems to be of influence on their quality of life.

Considering the problems with self-perception that have been proposed in NF1, this

observation in children self-reports is not surprising. However, why parents do not seem to

incorporate educational performance into HR-QOL ratings is not clear, in particular under the

consideration that learning disabilities are the most common complication of NF1 at pediatric

age.3 Possibly, this observation reflects a good acceptance of or resignation in school problems

in parents of children with NFL School problems, however, may influence QOL later on in life,

UJ\PTlcR-l 9 5

as they negatively influence educational level, job opportunities and SES. This has not been

investigated yet.

There are several potential limitations to our study. First, it should be noted that children

participated in the current study in the context of the baseline inventory of a larger study on

cognition and NFL 14 This may have resulted in a highly motivated group of families. The

children that participated in the current study (n=SS) did not differ significantly from the total

group eligible for the larger study (n=114) on age, sex, frequency of mental retardation, or

disease severity (all p2:0.3), indicating that they are representative for the total eligible group.

Second, the lack of Dutch reference values for the CHQ-PF of children older than 13 years may

limit the interpretation of our data. However, parental reports for children up to 13 years did

not significantly differ from reports of parents of children older than 13 years. Finally, our study

was not sufficiently empowered for subgroup-analysis on determinants of parent-child

discrepancies, or maternal or paternal effects on CHQ scores.

Conclusion

Parents report a profound impact of NF1 on physical, social, behavioral and emotional aspects

of HR-QOL. These findings underline the importance of multidisciplinary care of children with

NFl, encompassing not only physical but also social-emotional and behavioral assessment and

support. The substantial negative impact of behavioural problems on HR-QOL domains points

to an exciting potential opportunity to improve HR-QOL in children with NF1 by addressing

these behavioral problems. The fact that the children in our study report only minimal impact

on HR-QOL supports a deficit in self-perception in children with NF1, and emphasizes the

importance of cross-informant comparison of HR-QOL reports in order to obtain a

comprehensive overview of the impact of a disease on HR-QOL.36

list of abbreviations:

NFl, Neurofibromatosis type 1; CHQ, Child Healtb Questionnaire (PF, Parent Form; CF, Child Form); HR-QOL,

Healtb Related Quality of Life; QOL, Quality of Life; BP, Bodily Pain; GB, General Behavior; GH, General Healtb

Perceptions; PF, Physical Functioning; PT, Parental Time Impact; PE, Parental Emotional Impact; REB, Role

Functioning - Emotional /Behavioral; MH, Mental Healtb; SE, Self Esteem; FA, Family Activities; SES, Socio­

economical Status; TRF, Teacher's Report Form; ICC, Intraclass Correlation Coefficient; TACQOL, TNO-AZL Child

Quality of Life Questionnaire; ADHD, Attention Deficit Hyperactivity Disorder.

9 6 HEALTH RELATED QUALITY OF LIFE IN CHJLDERN WITH NEUROFIBROMATOSIS TYPE 1

References

1. Huson SM, Compston DA, Clark P, Harper PS. A genetic study of von Recklinghausen neurofibromatosis in south

east Wales. I. Prevalence, fitness, mutation rate, and effect of parental transmission on severity. J Med Genet

1989;26:704-11.

2. Ferner RE, Huson SM, Thomas N, Moss C, Willshaw H, Evans DG, et al. Guidelines for the diagnosis and

management of individuals with neurofibromatosis 1. J Med Genet 2007;44:81-8.

3. Ozonoff S. Cognitive impairment in neurofibromatosis type 1. Am J Med Genet 1999;89:45-52.

4. Krab LC, Aarsen FK, de Goede-Bolder A, Catsman-Berrevoets CE, Arts WF, Moll HA, et al. Impact of

Neurofibromatosis type 1 on School Performance. J Child Neurol2008;23(9), in press ( doi:10.1177 /0883073808316

366).

5. Johnson NS, Saal HM, Lovell AM, Schorry EK Social and emotional problems in children with neurofibromatosis

type 1: evidence and proposed interventions. J Pediatr 1999;134:767-72.

6. Dilts CV, Carey JC, Kircher JC, Hoffman RO, Creel D, Ward K, et al. Children and adolescents with

neurofibromatosis 1: a behavioral phenotype. J Dev Behav Pediatr 1996;17:229-39.

7. Barton B, North K Social skills of children with neurofibromatosis type 1. Dev Med Child Neurol2004;46:553-63.

8. Hyman SL, Shores A, North KN. The nature and frequency of cognitive deficits in children with neurofibromatosis

type 1. Neurology 2005;65:1037-44.

9. Descbeemaeker MJ, Ghesquiere P, Symons H, Fryns JP, Legius E. Behavioural, academic and neuropsychological

profile of normally gifted Neurofibromatosis type 1 children. J Intellect Disabil Res 2005;49:33-46.

10. Wolkenstein P, Zeller J, Revuz J, Ecosse E, Leplege A. Quality-of-life impairment in neurofibromatosis type 1: a

cross-sectional study of 128 cases. Arcb Dermatol2001;137:1421-5.

11. Page PZ, Page GP, Ecosse E, Korf BR, Leplege A, Wolkenstein P. Impact of neurofibromatosis 1 on Quality of

Life: a cross-sectional study of 176 American cases. AmJ Med Genet A 2006;140:1893-8.

12. Oostenbrink R, Spong K, de Goede-Bolder A, Landgraf JM, Raat H, Moll HA. Parental reports of health-related

quality of life in young children with neurofibromatosis type 1: influence of condition specific determinants. J Pediatr

2007;151:182-6, 6 e1-2.

13. Graf A, Landolt MA, Mori AC, Boltshauser E. Quality of life and psychological adjustment in children and

adolescents with neurofibromatosis type 1. J Pediatr 2006;149:348-53.

14. Krab LC, De Goede-Bolder A, Aarsen FK, Pluijm SM, Bouman MJ, VanderGeest JN, et al. Effect of simvastatin

on cognitive functioning in children with Neurofibromatosis type 1: a randomized controlled trial. JAMA

2008;300(3):287 -94

15. Landgraf J, Abetz L, JE. W. The CHQ user's manual. Boston: The Health Institute, New England Medical Center;

1996.

16. Raat H, Bonsel GJ, Essink-Bot ML, LandgrafJM, Gemke RJ. Reliability and validity of comprehensive health status

measures in children: The Child Health Questionnaire in relation to the Health Utilities Index. J Clin Epidemiol

2002;55:67-76.

17. Raat H, Landgraf JM, Bonsel GJ, Gemke RJ, Essink-Bot ML. Reliability and validity of the child health

questionnaire-child form (CHQ-CF87) in a Dutch adolescent population. Qual Life Res 2002;11:575-81.

18. Divisie Sociale en Ruimtelijke statistieken, Sector Ontwikkeling en ondersteuning. Standaard Beroepenclassificatie

1992. 2001 ed. Voorburg/Heerlen: Centraal Bureau voor de Statistiek.

19. Ablon J. Gender response to neurofibromatosis 1. Soc Sci Med 1996;42:99-109.

20. Friedman JM, Gutmann, D.H., Maccollio, M., Riccardi, V.M. Neurofibromatosis: Phenotype, Natural History, and

Pathogenesis. 3 ed. Baltimore and London: The Johns Hopkins University Press; 1999.

CR\PTER4 9 7

21. Achenbach TM. Integ<ative guide for the CBCL/4-8, YRS and TRF profiles: University of Vermont Department of

Psychiatry, Burlington, VT; 1991.

22. Raat H, Landg<af JM, Oostenbrink R, Moll HA, Essink-Bot ML. Reliability and validity of the Infant and Toddler

Quality of Life Questionnaire (ITQOL) in a general population and respiratory disease sample. Qual Life Res

2007;16:445-60.

23. Cohen J. Statistical power analysis for the behavioral sciences. New York: Academic Press; 1977.

24. Landis JR, Koch GG. The measurement of observer ag<eement for categorical data. Biometrics 1977;33:159-74.

25. Vogels T, Verrips GHW, Koopman H.M., Theunissen N.C.M., Fekkes M, Kamphuis RP. TACQOL manual: parent

form and child form. Leiden: Leiden Center for Child Health and Pediatrics LUMC-TNO; 2000.

26. Zoller ME, Rembeck B. A psychiatric 12-year follow-up of adult patients with neurofibromatosis type 1. J Psychiatr

Res 1999;33:63-8.

27. Barton B, North K. The self-concept of children and adolescents with neurofibromatosis type 1. Child Care Health

Dev 2007;33:401-8.

28. Sebold CD, Lovell A, Hopkin R, Noll R, Schorry E. Perception of disease severity in adolescents diagnosed with

neurofibromatosis type 1.J Adolesc Health 2004;35:297-302.

29. Heath NL, Glen T. Positive illusory bias and the self-protective hypothesis in children with learning disabilities. J

Clin Child Adolesc Psychol2005;34:272-81.

30. Hoza B, Pelham WE, Jr., Dobbs J, Owens JS, Pillow DR. Do boys with attention-deficit/hyperactivity disorder have

positive illusory self-concepts? J Abnorm Psychol2002;111:268-78.

31. White-Koning M, Arnaud C, Bourdet-Loubere S, Bazex H, Colver A, Grandjean H. Subjective quality of life in

children with intellectual impairment--how can it be assessed? Dev Med Child Neurol2005;47:281-5.

32. Klassen AF, Miller A, Fine S. Ag<eement between parent and child report of quality of life in children with attention­

deficit/hyperactivity disorder. Child Care Health Dev 2006;32:397 -406.

33. White-Koning M, Arnaud C, Dickinson HO, Thyen U, Bed.-ung E, Fauconnier J, et al. Determinants of child-parent

agreement in quality-of-life reports: a European study of children with cerebral palsy. Pediatrics 2007;120:e804-14.

34. von Rueden U, Gosch A, Rajmil L, Bisegger C, Ravens-Sieberer U. Socioeconomic determinants of health related

quality of life in childhood and adolescence: results from a European study. J Epidemiol Community Health

2006;60:130-5.

35. Mautner VF, Kluwe L, Thakker SD, Leark RA. Treatment of ADHD in neurofibromatosis type 1. Dev Med Child

Neurol2002;44:164-70.

36. Eiser C, Morse R. Can parents rate their child's health-related quality of life? Results of a systematic review. Qual Life

Res 2001;10:347-57.

9 8 HEALTH RElATED QUALITY OF LIFE TN CHTLDERN WJTH NEUROFIBROMATOSIS TYPE 1

Crl \PTicR 4 I 9 9

The rye/ink camera (left) and the prism adaptation setup (right).

Printed with permission.

Lianne C. Ktab, MSc43, Arja de Goede-Bolder, MD1, F emke K. Aarsen, MA 2,

Henriette A. Moll, MD PhD\ Chris 1. DeZeeuw, MD PhD3•4, Ype Elgersma, PhD3,

Jose.f N. van der Geest, PhD3• #

1 Department of Gener:.lf PediatJics, 2 Department of Pediatric New·ology, 3 Department of Neuroscience, Erasmus MC /Sophia Children's Hospital­NFJ CoRe Team (Cognitive Research Team), Rotterdam, The Netherlands.

4Netherlands Institute for New-oscience, Amstudam, The Netherlands

Submitted (2008)

Abstract

Purpose

Neurofibromatosis type 1 (NF1) is characterized by various neurocutaneous symptoms and

cognitive impairments. In addition, children with NF1 have frequently been reported to display

problems in fine and gross motor performance. However, their alleged deficits in motor learning

capacities have so far not been investigated systematically.

Methods

We investigated motor performance and motor learning in 70 children with NF1 and 19 healthy

age-matched controls (8-16 years) using various quantitative tests. We used the Beery

Developmental test for Visual-Motor Integration (Beery VMI) to assess fine motor performance

and visuo-motor integration controlled by mainly cerebral processing, and paradigms for

saccadic eye movement adaptation and prism-induced hand movement adaptation to assess

motor performance and motor learning capacities controlled by mainly cerebellar processing.

Results

NFl children scored significantly lower on the Beery VMI, showing problems in both visuo­

motor integration as well as in fine motor coordination. While no significant impairments were

observed in motor performance of either eye or arm movements, NFl children did show

deficits in motor learning during prism-induced hand movement adaptation. In contrast,

saccadic eye movement adaptation appeared not to be affected in NFL No correlation was

observed between scores on any of the thtee paradigms assessed.

Conclusions

Taken together, our results suggest that the motor problems of children with NF1 in daily life

may partly be related to deficits in motor learning. These behavioral deficits may be caused by

aberrations within specific regions of the cerebellum and cerebrum, but not by a ubiquitous

malfunctioning of these brain regions as a whole.

102 MOTOR LEARNING IN CHILDREN WITH NEUROFIBROMATOSIS TYPE 1

Introduction

Neurofibromatosis type 1 (NF1, incidence 1:3000)1 is an autosomal dominant disease caused by

a mutation in the gene for neurofibromin on chromosome 17q11.2. NF1 is characterized by a

variety of neurocutaneous symptoms and cognitive problems, the latter resulting in a lowered

mean IQ and a variety of school problems.2 In addition, many NF1 patients display impairments

in fine and/ or gross motor function, and over 40 percent of the NF1 children receive remedial

teaching to alleviate or improve motor performance.2 Fine motor problems are reported in areas

of fine motor coordination, fine motor speed, and steadiness.3, 4 One of the neuropsychological

tests consistently reported to be impaired in NF1 patients is the Beery Developmental test for

visual-motor integration (Beery VMIS), a test for fine motor coordination and the integration

between the visual-perceptual and motor abilities.2. 6-8 Gross motor problems observed in NF1

include hypotonia and problems with motor coordination, balance and gait.3. 7, 9

Although it is likely that the fine and gross motor problems in NF1 arise from deficits in a

network of brain areas, the cerebellum could be of particular interest in NFL The involvement

of this particular brain structure in NF1 is suggested by behavioral, radiological, and molecular

studies of NF1. First, although NF1 patients are not clearly ataxic, the frequently reported

clumsiness in movements 10· 11, could be related to deficits in the vermis, intermediate or lateral

zones of the cerebellum.12 Second, the cerebellum is one of the predominant sites for NF1

related hyperintensities visible on T2-weighted MR images, which have been related to

impairment of fine motor skills.3 Third, NF1 specifically seems to affect GABAergic neurons,B-

15 and the cerebellar GABA-agric Purkinje neurons are among the highest neurofibromin

expressing neurons in the brain,16, 17

The cerebellum plays an important role in motor performance, but also in motor learning,

which refers to the ability to continuously adapt movements to optimize performance, a task

which requires neuronal plasticity.18-24 The motor learning capacities of children with NF1 have

not been investigated so far. In the present study we quantitatively assessed motor performance

and motor learning in a large group of children with NFL

We assessed fine motor performance using the Beery VMI test, and cerebellar-mediated motor

performance and motor learning using tests on eye movement and hand movement control,

which are affected in patients with cerebellar deficits,23. 25-30 Performance and plasticity of

saccadic eye movements was examined in a saccade adaptation paradigm,zs which assesses the

CH,\PTER 5 103

gradual modification of the amplitude of saccadic eye movements induced by a systematic

change in the visual environment.19 Performance and plasticity of hand movement control was

assessed using prism adaptation, which refers to the modification of hand movement trajectories

in response to visual displacement of the environment induced by wearing prism goggles.31 We

hypothesized that motor learning capacities in children with NF1 are affected.

Methods

Subjects

70 children with NF1 (age 12.3 ± 2.5 years, 36 boys, 34 girls) and 19 healthy control children

(age 10.7 ± 2.1 years, 6 boys, 13 girls) participated in this study. Children with NF1 were

recruited from the patient group attending the NF1 outpatient clinic of the Erasmus MC -

Sophia Children's Hospital in Rotterdam. Some of these children participated in this study in the

context of a larger study of NF1 and cognition.2 Inclusion criteria were NF1 diagnosis according

to the criteria of the National Institutes of Health32 and informed consent from the parents and

from the children aged 12 years and older. Exclusion criteria were segmental NF1, pathology of

the CNS (other than asymptomatic gliomas), deafness, severely impaired vision, use of anti­

epileptics, inefficient production or comprehension of the Dutch language, and severe mental

retardation (IQ below 48). The control subjects were children of employees of the Erasmus MC

- Sophia Children's Hospital. The study was approved by the Medical Ethical Committee of the

Erasmus MC.

Procedure

Subjects participated in three tasks the Beery VMI, a Saccade Adaptation test and a Prism

Adaptation test.

Beery VMI- Visual Motor Integration

Fine motor coordination and visual motor integration was assessed with the Beery VMI task,S in

which children have to imitate or copy up to 30 geometric forms with increasing complexity

using paper and pencil. The test was stopped when a child makes more than two errors in a row.

Copying errors were marked if they reflected problems in fine motor coordination, rather than a

pure visual-spatial problems. The task is specifically designed for children and takes about 10

minutes. Beery VMI scores were standardized for age and sex using normative data for the

general population.s Differences between the two groups were assessed using Kolmogorov­

Smirnov tests.

104 M:OTORLEARNING IN CI-ULDREN WITH NEUROFJBRO!vlATOSlS TYPE 1

Saccade Adaptation - Eye movement control

Performance and plasticity of saccadic eye movements was assessed in a classical backward

saccade adaptation paradigm.JD, 33 Subjects were seated 70 em in front of a 21-inch computer

screen. This experiment took place in complete darkness. A red filter covered the computer

screen to eliminate all light emitted by the monitor other than the visual stimuli. Binocular eye

position was recorded using infrared video-oculography (EyeLink 2.04, SensoMotoric

Instruments, Berlin, Germany) at a sample rate of 250 Hz.3D. 34 Eye position was calibrated with

the built-in 9 points calibration routine. A chin rest ensured a stable position of the head and

head movements were monitored using the built-in head-tracking camera.

The saccade adaptation paradigm consisted of three distinct phases: 20 baseline trials, followed

by 100 adaptation trials, and 20 extinction trials. In all phases the subjects were instructed to

look at a single red dot (0.5 degrees of visual angle in diameter) that jumped from left to right.

Each trial started with the dot being displayed at 7.5 degrees of visual angle on the left side from

the center of the screen. After fixation the dot was removed on the left and subsequently

displayed 7.5 degrees from the center on the right side of the screen, evoking a primary saccadic

eye movement from left to right with a target amplitude of 15 degrees. In the baseline and the

extinction trials the dot remained on the right side of the screen for 1.5 seconds after which the

next trial was started. In the adaptation trials the dot on the right stepped 3 degrees to the left,

i.e., 20 percent of the initial target amplitude bach.-wards, during the saccadic eye movement

toward it.

The amplitude of the primary saccade was determined for each of the 140 trials. Trials were

discarded when the primary saccade did not start on the left side, was not directed toward the

target on the right, or had an amplitude of less than 8 degrees. For all trials, the saccadic Gain

was defined as the amplitude of the primary saccade divided by the target amplitude (15

degrees), so that a gain of 1 reflects a saccade that lands directly on target.

For each subject, the Baseline Gain was calculated as the average of the gains of the primary

saccades made in the 20 baseline trials, and the Adapted Gain as the average of the gains of the

last 20 trials in the adaptation phase. For each subject, the saccadic Gain Change was calculated

as the difference between Adapted Gain and Baseline Gain. Saccadic Variability in the baseline

and adapted phase was defined as the within-subject standard deviation of the primary saccadic

gains in these phases.

CH \PTER 5 105

We defined saccade adaptation significant in an individual, when the Gain Change was larger

than the average minus 1 standard deviation of the control group combined with a significant

(p<0.01) difference between Baseline and Adapted Gains. Participants were excluded from

further analyses if the average baseline gain or the baseline saccadic variability was an outlier or

extreme value (value more than 1.5 interquartile ranges below the 25th or above the 75th

percentile).

Prism Adaptation - Hand movement control

The performance and plasticity of hand movement coordination was determined in a prism

adaptation experiment.zB Subjects were seated in front of a digitizing tablet (Ultrapad A2,

WACOM Technologies Corporation, Vancouver, WA, USA). The target (a small cartoon

picture) was projected from above on a see-through mirror, so that it seemed to be positioned

on the tablet 20 em straight ahead of the subject, while the hand was also visible. Visual

feedback of hand position could be blocked by putting an opaque plate below the mirror, so

that the target was still visible through the mirror but the hand below the mirror could no longer

be seen (see van der Geest et al.ZB for details of the setup).

The experiment consisted of four phases. In all phases subjects had to move the pen a number

of times from a starting position at the left bottom of the tablet (17 em from the center) towards

the position of target over a movement distance of 26 em with an angle of 50 degrees. In the

practice phase (phase 1) the subject had to move the pen towards the target 10 times while they

could see their hand (visual feedback). In the pre-adaptation phase (2) the subject had to move

the pen 10 times without visual feedback. In the adaptation phase (3) the subject wore prism

glasses that shifted the visual world 10 degrees to the right. Subjects had to move the pen 30

times to the target and two additional practice-targets positioned about 17cm to the left and

right of the original target. In this phase they could see their hand again, so that the position of

the hand and target could be visually aligned. Before the post-adaptation phase (4), the glasses

were removed and subjects had to move the pen 10 times without visual feedback.

The end-position of each hand movement across the tablet toward the target was marked

manually. The movement angle (in degrees) and the movement distance (in em) was calculated

from the straight line between start- and end-position of the movement. For each subject, the

averages and standard deviations of the movement angles and distances in the baseline phase,

the pre-adaptation and the post-adaptation phase were determined. To assess the effect of

106 MOTOR LEARNING IN CHILDREN W1TH NEUROFIBRO.MATOSIS TYPE 1

wearing prism glasses (prism adaptation, also called the after-effect24), the change in average

movement angle (Angle Change) between the pre- and post-adaptation phase was calculated.

We defined prism adaptation significant in an individual, if the change in movement angle was

larger than mean minus 1 standard deviation of the controls and the difference between the

average pre- and post adaptation angle was significant (p<O.Ol). Subjects had to hold the pen in

their dominant right hand. Therefore, seven NF1 children and one control child who were left­

handed were not eligible for this task.

To assess motor performance we compared the Beery VMI scores, baseline Saccadic Variability

and variability in hand movement angle in the pre-adaptation phase between the two groups. To

assess motor learning we compared changes in saccadic gain and changes in movement angles

between the two groups. Statistical differences were assessed non-parametrically using Mann

Whitney, Chi-Square, and Kolmogorov-Smirnov tests. Spearman correlations between Beery

VMI scores and the motor performance and motor learning measures, and age were calculated.

Results

BeeryVMI

Beery VMI scores (Figure 1A) were significantly lower in the NFl group (84 ± 13, n=70) than

in the control group (102 ± 14, n=19, absolute extreme difference = .67, Z=2.58, p<0.001).

Control children completed more items than NF1 children (on average 22.3 ± 2.0 versus 19.6 ±

3.9, absolute extreme difference .371, Z = 1.44, p < 0.05) before the test was stopped. In the

copying errors made in the NFl group, but also in the control group, visual-spatial problems as

well as problems in fine motor coordination were observed (see figure 1B). In the NF1 group

about 50% of the copying errors were related to problems of fine motor coordination, rather

than to pure visual-spatial problems, which was, however, not significantly different from

controls.

Saccade adaptation

70 children with NF1 and 19 control children performed the saccade adaptation test. 17 NF1

children and seven controls were excluded from analysis because of technical failures, including

eye tracking difficulties, making too large head movements and making too few saccades for

proper analysis.

CI-L\PTFR 5 107

A 60

I~~~~trol. 50

I

1

~ 40

Q)

:c 30 ::l 1/J

::R 0 20

Beery Score

Figure 1: Beery VMI.

Panel A shows the distribution of Beery VMI scores in 70 NF1 children and 19 age-matched controls; panel B shows

examples of Beery VMI performance of 2 NF1 children and of 2 age-matched control children with around average Beery

VMI score for their respective groups. Items illustrating pure motor problems in these 2 NF children were selected.

Performance on item 2 is shown for a male NF1 child (age 14.6y, score 79) and a male control (age 14.3y, score 97). The

NF1 child drew an unsteady line, had a weak pencil stroke, and there was an indication of a very discrete tremor. Item 16

is shown for a male NFl child (age 10.9y, score 84) and a female control (age 10.5y, score 103). The NF1 child shows a

general delay in fine motor development and performs around developmental age 5.4y on this item.s Note the slip of the

pencil at the end of the movement.

We observed no differences in baseline saccadic performance between the remaining 53

children with NF1 (28 boys, 25 girls, 12.6 ± 2.3 years) and controls (2 boys, 9 girls, 10.8 ± 2.1

years). Specifically, the number of correct primary saccades in the 140 trials (122 ± 9 for NF1

vs. 124 ± 9 for controls, p=0.5), the baseline Saccadic Gains (0.91 ± 0.08 versus 0.93 ± 0.04,

p=0.4) and baseline Saccadic Variability (0.10 ± 0.04 versus 0.08 ± 0.02, p=0.2) did not differ

between the two groups (figure 2A).

Saccadic adaptation was also not different in NFl children compared to controls (figure 2B)

with respect to the size of the Adapted Gains (0.78 ± 0.10 for NFl vs. 0.78 ± 0.10 for controls,

p=0.9) and the adapted Saccade Variability (0.09 ± 0.03 versus 0.09 ± 0.02, p=l.O). The saccadic

Gain Change between baseline and the end of the adaptation phase was also not significantly

different between NF1 children and controls (0.12 ± 0.08 versus 0.15 ± 0.09, p=0.3). The

proportion of subjects with a significant Gain Change (Gain Change >0.06 as derived from the

control group) was the same in the two groups (29 out of 53 NF1 children (55%) versus 7 out

of 11 controls (64%), x2=0.294, p=0.6). The distribution of individual Gain Changes was also

not significantly different (absolute extreme difference = 0.235, 2=0.709, p=0.7, figure 2C).

108 MOTOR LEARNING IN CHILDREN WITH NEUROFJBROI\iATOSlS TYPE 1

The gain changes were not related to age in NF1 children or controls (R=0.04, p=0.8 for NF1

and R= -0.13, p=0.7 for controls).

A 0.4

.= 0.3

"'' (!)

. s ~0.2 :0: I'll ·c

"'' > 0.1

8.6

....

" .. .,. . . ;l.j ..... . ···~ • • • •

•

• Controls •NF1

•

B 1.4-

1.2

• • • .,,. ...... . ....... ... ,.

\. ,._.,. • •• ..

•

•Controts •NF1

0.8 1 1.2 1.4 0.8 1 1.2

Average Gain Baseline Gain

c --Controls -NF1

01~'------------~----------==~--------0.1 0 0.1 0.2 0.3 0.4

Change in Angle

Figure 2: Saccade adaptation.

1.4

Panel A shows the variability versus the average of the baseline saccadic gains of 53 NF1 children and 11 age-matched

controls; each dot represents one individual subject. Panel B shows the adapted gain versus the baseline gain for these

children; the oblique line is the unity line. Panel C shows the cumulative distribution of the Gain Changes in the NF1 and

control groups. The vertical line (at Gain Change = 0.06) indicates the cur-off for point significant saccade adaptation.

Prism adaptation

63 right-handed NF1 children and 18 right-handed control children were eligible for the prism

adaptation task. Seven NF1 children were excluded from analysis because of technical problems

including not understanding or adhering to task instructions. All remaining 56 children with

NF1 (29 boys, 27 girls, 12.3 ± 2.4y) and 18 controls (5 boys, 13 girls, 10.6 ± 2.2y) were able to

make accurate goal-directed hand movements towards the target. As expected, for both groups

0-L\PTER 5 109

the movement angle was about 50 degrees and the movement distance was about 26 em when

children could align their hand visually with the target in the baseline phase. Without visual

feedback (pre-adaptation phase) both groups became less accurate but no difference between

the two groups was observed (movement angle: 56.8 ± 3.2 degrees in NF1 vs. 55.6 ± 2.8

degrees in controls, p=0.2; distance: 24.0 ± 2.3 em in NF1 vs. 24.0 ± 2.0 em in controls, p=0.9,

see figure 3A).

A 5

Oi Q.l 4 E. Q.l

"5:1 3 c <(

.!: >- 2 = :0 . ~ ii >

~5

• Controls

•NF1

• • • ' • • •• • .. , ...

• •

..

• . "' .. .. •• •• .. .; .. :"' • ... - • •• .. .,,, . .. .. .,. .,

" ••• •

.. •

• Controls •NF1

50 55 60 65 50 55 60 65 70

Average Ang:le (deg) Pre-adaptation Angle (deg)

c - Controls

NF1

0 ~-2----0-----2~--4--~s~~s======

Change in Angle (deg) Figure 3: Prism adaptation

Panel A shows the variability versus the average hand movement angle of 56 NFl children and 18 healthy controls in the

pre-adaptation condition (without visual feedback of the hand); each dot represents one individual subject. Panel B shows

the average angle of the arm movements in the post-adaptation phase versus the average angle of the arm movements in

the pre-adaptation phase for these children; the oblique line is the unity line. Panel C shows the cumulative distribution of

the changes in average movement angles between the pre- and post adaptation phases in the NFl and the control group.

The vertical line indicates the cur-off point for significant prism adaptation (2.93 degrees).

After wearing prism goggles with visual feedback in the adaptation phase, the average

movements in the post-adaptation phase did not differ between the groups (movement angle:

110 MOTOR LEARNING IN CHILDREN WTTH NEUROF!BROIVL'\TOSIS T'iPE 1

59.9 ± 3.6 degrees in NF1 vs. 60.1 ± 2.3 degrees in controls, p=0.8; distance: 24.0 ± 2.3 em in

NF1 vs. 23.7 ± 1.9 em in controls, p=0.7). However, the changes in movement angle between

the pre-adaptation and post-adaptation phase induced by the prism goggles was significantly

smaller in NF1 children than in controls (3.1 ± 3.0 vs. 4.5 ± 1.6 degrees, respectively, p=0.03,

see figure 3B).

As can be seen in figure 3B, some NF1 children did show a significant prism adaptation

(Change in angle> 2.9 degrees, as derived from the control group, with p<0.01). However, the

proportion of subjects with a significant adaptation tended to be smaller in the NF1 group (28

out of 56 NF1 children (50%)) than in the control group (13 out of 18 controls (72%), x2=2.72,

p=0.1). The difference in distributions of the changes in hand movement angles between the

two groups was marginally significant (absolute extreme difference = 0.375, Z=1.32, p=0.06, see

figure 3C).

Age was not related to performance on prism adaptation in children with NF1 or controls

(R=0.08, p=0.6 for NF1 and R= -0.32, p=0.2 for controls). We did not observe any correlations

between Beery VMI scores and the outcomes of the motor tests in NF1 children and controls

(all p>0.5).

Discussion

In the present study, motor performance and motor learning capacities of children with NF1

was assessed using the Beery VMI, saccade adaptation and prism adaptation tasks, and

compared to healthy age-matched controls. As expected, children with NF1 show lower scores

on the Beery VMI task. In addition, the adaptation of hand movements to prism goggles was

reduced in NF1 children. However, saccadic performance and plasticity were not affected, as

well as the performance of goal directed hand movements.

As expected, NF1 children scored significantly lower on the Beery VMI task than controls. We

observed the typical visual-spatial problems, which are in line with the poorly developed visual­

spatial skills of children with NF1 (reviewed by Ozonof£)35, but we found that some copying

errors were related to problems fine motor coordination. Impairments in the Beery VMI can

indicate problems in a variety of brain areas, including the right hemisphere, the primary motor

cortex of the dominant hand, the cerebellum, subcortical nuclei, and/ or the corpus callosum.5

CH \PTER5 111

Baseline saccadic accuracy and the ability to adapt saccadic eye movements in NF1 appeared

normaL In the present study, 55% of the NF1 and 64% of the control children were able to

modify the amplitudes of their saccades in a classical saccade adaptation paradigm. These

percentages are in good agreement with the 66% found in a group of 39 healthy children.36 The

saccadic oculomotor system comprises a network of brain areas, including the cerebellum and

the parietal and frontal cortex, the basal ganglia, the superior colliculus and the brainstem.37 The

cerebellar oculomotor vermis is critically involved in maintaining a high saccadic accuracy.19·21

Impairments in saccadic latencies and directions have been observed previously in a small group

of 10 children with NF1, which were postulated to reflect involvement of a broad (cortical)

network of brain areas involved in saccadic controJ.3s. Our results suggest that the oculomotor

vermis of the cerebellum is less likely to be part of that potentially deficient network in NFL

Prism adaptation seems to be impaired in children with NFL Although both groups show a

significant change in the angle of the hand movements after prism glass displacement, the

average degree of adaptation is significantly smaller in NF1 children. Furthermore, the variability

in performance between NF1 children is larger than between control children and fewer NFl

children show a significant prism-induced after-effect. A subgroup of NF1 children could adapt

their hand movements quite adequately, whereas others did not adapt at all. This seems to be in

line with the large variability in clinical and cognitive characteristics between patients with

NF1.39 Impaired prism adaptation could result from problems in the anterior and caudal

posterior lobe of the cerebellar cortex (including Cl - C3 zones), but also upon other motor

areas such as the ventral premotor cortex and the posterior parietal cortex, which is involved in

visually directed movements.24,40

A potential limitation of our study is that a few NF1 children did not perform properly in the

saccade and prism adaptation tasks, for instance by making large head movements, or by not

making goal directed hand movements, which excluded them from further analysis. This was

likely due to a short attention span or inability to understand the instructions completely.

Although these tests have been administered successfully in children with and without mental

retardation zs. 30· 36, it could be that the attention and cognitive deficits in these excluded children

is related to worse motor performance. Their exclusion may have led to an overestimation of

the performance of NF1 children on this task.

Taken together, our findings suggest that the motor problems displayed by children with NF1

may partially be related to deficits in plasticity of motor control.41 However, our results suggest

112 MOTOR LEARNING IN CHILDREN WITH NEUROFIBRO!v1ATOSIS TYPE 1

that these deficits do not arise from a single brain region as a whole, such as the cerebellum.

Rather, specific parts of the cerebellum and cerebrum may be selectively affected. Future

research is necessary to unravel the neuronal basis of motor problems in patients with NFL

Acknowledgements

We are very grateful to the patients and their parents for their participation. Also, we thank E.

Barendse, S.M. Goorden, A.C. Gaemers, P. Plak, R. Wierenga, and M.J. Bouman for their help

in collecting the data. We appreciate the support of all other participants of the NF1 clinical

workgroup and the NF1 CoRe (Cognitive Research) team of the Erasmus MC - Sophia

Children's Hospital Rotterdam, the Netherlands. This study was supported by the

Hersenstichting Nederland, the Sophia Foundation for Medical Research, the Prinses Beatrix

Fonds, and by a donation from the Dutch Neurofibromatosis Foundation.

Documentation of Author Roles

This study was conceived by JNvdG, CidZ, and YE, and funding was obtained by CidZ and

YE. The study was organized by LCK, JNvdG, AdGB, HAM, and YE. Data collection was

performed by LCK, AdGB, FKA, and JNvdG. Statistical analyses were designed and executed

by LCK, JNvdG and FKA, and, in addition, reviewed by YE. The first draft of the manuscript

was written by LCK and JNvdG, and critically reviewed by all authors.

CR\PTER5 113

References

1. Huson SM, Harper PS, Compston DA. Von Recklinghausen neurofibromatosis. A clinical and population study in

south-east Wales. Brain 1988;111 ( Pt 6):1355-1381.

2. Krab LC, Aarsen FK, de Goede-Bolder A, et al. Impact of Neurofibromatosis type 1 on School Performance. J

Child Neurol2008;23(9), in press ( doi:10.1177 /0883073808316 366) ..

3. Feldmann R, Denecke J, Grenzebach M, Schuierer G, Weglage J. Neurofibromatosis type 1: motor and cognitive

function and T2-weighted MRI hyperintensities. Neurology 2003;61(12):1725-1728.

4. Moore BD, Slopis JM, Schomer D, Jackson EF, Levy BM. Neuropsychological significance of areas of high signal

intensity on brain MRis of children with neurofibromatosis. Neurology 1996;46(6):1660-1668.

5. Beery KE, Beery NA. The Beery-Buktenica Developmental Test of Visual-Motor Integration, Fifth Edition.

Minneapolis NCS Pearson Inc., 2004.

6. Descheemaeker MJ, Ghesquiere P, Symons H, Fryns JP, Legius E. Behavioural, academic and neuropsychological

profile of normally gifted Neurofibromatosis type 1 children. J Intellect Disabil Res 2005;49(Pt 1):33-46.

7. North K, Joy P, Yuille D, et al. Specific learning disability in children with neurofibromatosis type 1: significance of

MRI abnormalities. Neurology 1994;44(5):878-883.

8. Dilts CV, Carey JC, Kircher JC, et al. Children and adolescents with neurofibromatosis 1: a behavioral phenotype. J

Dev Behav Pediatr 1996;17(4):229-239.

9. Chapman CA, Waber DP, Bassett N, Urian DK, Korf BR. Neurobehavioral profiles of children with

neurofibromatosis 1 referred for learning disabilities are sex-specific. Am J Med Genet 1996;67(2):127 -132.

10. Cnossen Ivlli, de Goede-Bolder A, van den Broek KM, et al. A prospective 10 year follow up study of patients with

neurofibromatosis type 1. Arch Dis Child 1998;78(5):408-412.

11. Korf BR. Clinical features and pathobiology of neurofibromatosis 1. J Child Neurol2002;17(8):573-577; discussion

602-574, 646-551.

12. Gazzaniga MS, Ivry RB, Mangun GR. Cognitive Neuroscience - The biology of the mind. In: WW Norton &

Company, 1998:Chapter 10.

13. Costa RM, Federov NB, Kogan JH, et al. Mechanism for the learning deficits in a mouse model of

neurofibromatosis type 1. Nature 2002;415(6871):526-530.

14. Dasgupta B, Gutmann DH. Neurofibromatosis 1: closing the GAP between mice and men. Curr Opin Genet Dev

2003;13(1):20-27.

15. Gutmann DH, Collins FS. The neurofibromatosis type 1 gene and its protein product, neurofibromin. Neuron

1993;1 0(3):335-343.

16. Nordlund ML, Rizvi TA, Brannan CI, Ramer N. Neurofibromin expression and astrogliosis in neurofibromatosis

(type 1) brains. J Neuropathol Exp Neurol1995;54(4):588-600.

17. Gutmann DH, Geist RT, Wright DE, Snider WD. Expression of the neurofibromatosis 1 (NF1) isoforms in

developing and adult rat tissues. Cell Growth Differ 1995;6(3):315-323.

18. Koekkoek SK, Hulscher HC, Dortland BR, et al. Cerebellar LTD and learning-dependent timing of conditioned

eyelid responses. Science 2003;301(5640):1736-1739.

19. Hopp JJ, Fuchs AF. The characteristics and neuronal substrate of saccadic eye movement plasticity. Prog Neurobiol

2004;72(1):27-53.

20. Barash S, Melikyan A, Sivakov A, Zhang M, Glickstein M, Thier P. Saccadic dysmetria and adaptation after lesions of

the cerebellar cortex.] Neurosci 1999;19(24):10931-10939.

21. Desmurget M, Pelisson D, Grethe JS, et al Functional adaptation of reactive saccades in humans: a PET study. Exp

Brain Res 2000;132(2):243-259.

114 MOTORJ..EARNING IN CHILDREN WITH NEUROFIBROIVLATOSIS TI:l'E 1

22. Baizer JS, Kralj-Hans I, Glickstein M. Cerebellar lesions and prism adaptation in macaque monkeys. J Neurophysiol

1999;81 (4):1960-1965.

23. Pisella L, Rossetti Y, Michel C, et al. Ipsidirectional impairment of prism adaptation after unilateral lesion of anterior

cerebellum. Neurology 2005;65(1):150-152.

24. Clower DM, HoffmanJM, Vutaw JR, Faber TL, Woods RP, Alexander GE. Role of posterior parietal cortex in the

recalibration of visually guided reaching. Nature 1996;383(6601):618-621.

25. Martin TA, Keating JG, Goodkin HP, Bastian AJ, Thach WT. Throwing while looking through prisms. I. Focal

olivocerebellar lesions impair adaptation. Brain 1996;119 ( Pt 4):1183-1198.

26. Pisella L, Michel C, Grea H, Tilikete C, Vighetto A, Rossetti Y. Preserved prism adaptation in bilateral optic ataxia:

strategic versus adaptive reaction to prisms. E"-"P Brain Res 2004;156(4):399-408.

27. Fernandez-Ruiz J, Velasquez-Perez L, Diaz R, et al. Prism adaptation in spinocerebellar ataxia type 2.

Neuropsychologia 2007;45(12):2692-2698.

28. Van der Geest JN, Lagers-van Haselen GC, van Hagen JM, et al. Visual depth processing in Williarns-Beuren

syndrome. E"-"P Brain Res 2005;166(2):200-209.

29. Straube A, Deubel H, Ditterich J, Eggert T. Cerebellar lesions impair rapid saccade amplitude adaptation. Neurology

2001;57(11):2105-2108.

30. van der Geest JN, Haselen GC, Frens MA. Saccade adaptation in Williams-Beuren Syndrome. Invest Ophthalmol

Vis Sci 2006;47(4):1464-1468.

31. Redding GM, Rossetti Y, Wallace B. Applications of prism adaptation: a tutorial in theory and method. Neurosci

Biobehav Rev 2005;29(3):431-444.

32. NIH. National Institutes of Health Consensus Development Conference Statement: neurofibromatosis. Bethesda,

Md., USA,July 13-15,1987. Neurofibromatosis 1988;1(3):172-178.

33. Frens JvlA, van Opstal AJ. Transfer of short-term adaptation in human saccadic eye movements. Exp Brain Res

1994;100(2):293-306.

34. van der Geest JN, Frens MA. Recording eye movements with video-oculography and scleral search coils: a direct

comparison of two methods. J N eurosci Methods 2002; 114(2): 185-19 5.

35. Ozonoff S. Cognitive impairment in neurofibromatosis type 1. Am J Med Genet 1999;89(1) :45-52.

36. Salman MS, Sharpe JA, Eizenman M, et a!. Saccadic adaptation in children. J Child Neural 2006;21 (12):1 025-1031.

37. Robinson FR, Fuchs AF. The role of the cerebellum in voluntary eye movements. Annu Rev Neurosci 2001;24:981-

1004.

38. Lasker AG, Denckla MB, Zee DS. Ocular motor behavior of children with neurofibromatosis 1. J Child Neural

2003;18(5):348-355.

39. Easton DF, Ponder MA, Huson SM, Ponder BA. An analysis of varistion in expression of neurofibromatosis (NP)

type 1 (NF1): evidencefor modifYing genes. Am J Hum Genet 1993;53(2):305-313.

40. Kurata K, Hoshi E. Reacquisition deficits in prism adaptation after muscimol microinjection into the ventral

premotor cortex of monkeys. J Neurophysiol1999;81(4):1927-1938.

41. DeZeeuw CI, Yeo CH. Time and tide in cerebellar memory formation. Curt Opin Neurobiol2005;15(6):667-674.

OJ \PTLR 5 115

S.J.P.M. vanEngele11l, L.C. KtabZ,J, H.A. Molf\ A. de Goede·Bolder2, S.M.F. Pluijm\ C.E. Catsman-Berrevoets5, Y. Elgersma3, M.H. Lequin1

1 Department of Pediatric Radiology, 2 Gener~ll Paedi~ltdcs, 3 Neuroscience, 4 Public H e~llth, and -'Pediatric Neurology, The NFJ CoRe Team (Cognitilre Research Team), Erasmus MC-Sophia

Children's Hospital, Rotterdam, the Netherlands.

American Journal of Neuroradiology 2008;29(4): 816-22

Abstract

Background and Purpose:

Hyperintensities on T2-weighted images are seen in the brains of most patients with

Neurofibromatosis Type I (NFl), but the origin of these unidentified bright objects (UBOs)

remains obscure. In the current study, we examined the diffusion characteristics of brain tissue

in children with NFI to test the hypothesis that a microstructural abnormality is present in NFL

Materials and Methods:

Diffusion tensor imaging (DTI) was performed in 50 children with NFl and 8 controls. Circular

regions of interest were manually placed in 7 standardized locations in both hemispheres,

including UBO sites. Apparent diffusion coefficients (AD C), fractional anisotropy (FA) and

axial anisotropy (Am) were used to differentiate quantitatively between healthy and disordered

brain matter. Differences in eigenvalues (A.l, A.2, A.3) were determined to examine parenchymal

integrity.

Results:

We found higher ADC values for UBOs than for normal-appearing sites (p < 0.01), and higher

ADC values for normal-appearing sites than for controls (p < 0.04 in 5 of 7 regions). In most

regions, we found no differences in FA or Am. Eigenvalues A.2 and A.3 were higher at UBO sites

than in normal-appearing sites (p < 0.04).

Conclusion:

With ADC, it was possible to differentiate quantitatively between normal- and abnormal­

appearing brain matter in NFl and also between normal appearing brain matter in NFl and

healthy brain matter in controls, indicating subtle pathologic damage, disrupting the tissue

microstructure in the NFl brain. Higher diffusivity for A.l, A.2 and A.3 indicates that this

disturbance of microstructure is caused by accumulation of fluid or vacuolation.

118 QUANTITATIVE DII'FERENTIATION BETWEEN HEALTHY AND DISORDERED BRAIN

MATTER IN NEUROFIBROMATOSIS TYPE 1 PATIENTS USING DIFFUSION TENSOR

llVIAGlNG

Introduction

Hyperintensities on T2-weighted images are seen in the brains of most patients with

Neurofibromatosis Type I (NF1). Although many imaging techniques have been used to assess

these unidentified bright objects (UBOs), their origin remains obscure.1-4 The only pathologic

study performed in NF1 so far revealed intramyelinic vacuolar changes or spongiotic

myelinopathy that correlated with the hyperintensities found on T2-weighted images.s In

addition to conventional MR imaging, several studies have used diffusion-weighted-imaging

(DWI) with assessment of apparent diffusion coefficients (ADCs), to gain information on

UBOs that cannot be assessed by inspection of conventional images alone. On the basis of high

ADC values, a widespread myelin disorder was suggested to be present in patients with NF1.6-9

However, ADC reflects only the magnitude of the diffusion of water molecules. Although high

ADC values might suggest increased water content of the brain, with ADC alone, it is not

possible to examine the microstructural integrity of the parenchyma. Diffusion tensor imaging

(DTI), which measures the degree and direction of molecular diffusivity, is able to detect white

matter abnormalities and characterize them in terms of white matter fiber integrity.1°· 11

DTI generates a diffusion tensor matrix from a series of DWis. By matrix diagonalization, the 3

eigenvalues A.1, 1..2 and 1..3 can be calculated. 1..1 has the largest value and reflects the diffusivity

parallel to a structure; 1..2 and 1..3 are the middle and smallest eigenvalues, respectively; and their

average represents the diffusivity perpendicular to a structure. Various anisotropy indexes

(fractional anisotropy [FA], axial inisotropy [Am]) can be calculated by using the eigenvalues.

They describe the ratio of the eigenvalues and are scaled from zero (isotropic) to 1 (anisotropic)

and reflect the microstructure of white matter tracts. Looking at the eigenvalues themselves

enables specific assessment of myelin integrity, as distinct from axonal integrityJ2. 13 Recently, a

DTI study on adult patients with NFl revealed higher ADC and lower FA values.

Unfortunately, changes in eigenvalues were not reported.14

The purpose of this study was to examine the diffusion characteristics of brain tissue in children

with NF1 by means of DTI and to test the hypothesis that a microstructural abnormality is

present in NFL We tested this in three ways: First, by assessing ADC and indexes of anisotropic

diffusion, we tried to differentiate quantitatively between normal- and abnormal-appearing brain

tissue in children with NFL Second, we examined the normal-appearing parenchyma in NF1 to

see if it is different from parenchyma in healthy controls. Third, we looked at parenchymal

integrity at UBO sites and normal appearing sites by assessment of the eigenvalues. In addition,

CH.\PTER 6 119

because T2-weighted hyperintensities in the hippocampus have been suggested to have a

different pathogenetic basis from classic UBOs,lS we paid special attention to the diffusion

characteristics of the hippocampal hyperintensities to see if they were different from those in

other regions of interest.

Material and Methods

Subjects

Data for this study were obtained in the context of a larger study on NF1 and cognitive

functioning. AJl participants were recruited from the multidisciplinary NF1 outpatient clinic of

the hospital. Inclusion criteria were the following: age 8 to 1 7 years, NF1 diagnosis according to

the criteria of the National Institutes of Health16 and informed consent from parents and

children older than 12 years of age. Exclusion criteria were: segmental NF1 (because brain

involvement is not certain in these patients), pathology of the central nervous system (CNS)

(other than asymptomatic gliomas), deafness, severe impaired vision, use of antiepileptics,

inefficient production and comprehension of the Dutch language, and severe mental retardation

(intelligence quotient <48).

One hundred twenty-six Children fulfilled age criterion. Twelve children were excluded on the

basis of possible segmental NF1 (n = 3), use of antiepileptics (n = 3), hydrocephalus (n = 3),

severe mental retardation (n = 1) and inefficient production and comprehension of the Dutch

language (n = 2). The remaining 114 children were invited to participate in the larger study, of

which 62 consented. The study was approved by the medical ethics committee of our

institution. A total of 50 out of 62 children that participated in the larger study consented to :MR.

imaging (21 girls, mean age 12.2 years; range: 8.1-15.7 years, and 29 boys, mean age 12.3 years;

range: 8.0-16.5 years).

Image acquisition

:MR. anatomic imaging with DTI was performed using a l.ST system (EchoSpeed; GE

Healthcare, Milwaukee, Wis) and a dedicated 8-channel head coil. DTI data were acquired by

using a multi-repetition single-shot echo-planar sequence with a section thickness of 3 mm with

no gap. The DTI images were obtained in 25 gradient directions with a sensitivity of b = 1000

s/mm2, TR = 15000 ms, TE = 82.1 ms, 1 average, FOV of 240 x 240 mm2, a matrix of 128 x

128 resulting in a voxel size of 1.8 x 1.8 x 3.0 mm3. Acquisition time was 5:28 minutes with a

total of 53 sections to cover the entire brain.

120 QlJANTITATIVE DIFFERENTIATION BETWEEN HEALTHY AND DISORDERED BRAIN

1\iA.TTER IN NEUROFIBRO~·L'I.TOSJS TYPE 1 PATIENTS USING DIFFUSION TENSOR IMAGING

Data collection

All images were analyzed by visual inspection by an experienced pediatric neuroradiologist to

exclude CNS tumors. Hyperintense lesions on T2-weighted, fluid-attenuated inversion recovery

(FLAIR) and diffusion images (b = 0 s/mm2) were classified as UBOs, when no hyperintense

lesion was present, the area was scored as normal-appearing site.

For quantitative data analysis, ADC, FA, Am, and eigenvalues were used. FA measures the

fraction of the magnitude of the diffusion tensor that can be ascribed to anisotropic diffusion.17

Am reflects the shape of the diffusion ellipsoid.1B ADC, FA, and eigenvalue maps were

reconstructed by using commercially available software (Functool3.1.23, Advanced Workstation

4.1; GE Healthcare). Circular regions of interest of specific sizes were manually placed in 7

predetermined anatomic locations in both hemispheres: the cerebral peduncle (CP), cerebellar

white matter (CWl\1), hippocampus (HI), thalamus (fH), globus pallidus (GP), and frontal

(FWl\1) and parieto-occipital (POWl\1) white matter. The size of the regions of interest was 100

mm2 for CWM, 70 mm2 for CP, 170 mm2 for HI, POWM and FWM, and 130 mm2 for GP and

TH, according to the method of Alkanet al (Fig 1).8 All regions of interest placement was done

on the b = 0 s/mm2 images because anatomic detail was better on these images than on the

computed maps, ensuring anatomic precision. Regions of interest were automatically

superimposed on the functional maps by the software that was used in this study. ADC maps

were used to exclude CSF from measurements to minimize overestimation of the ADC values.

FA maps were used when placing regions of interest in the G P and TH region to avoid as much

as possible the involvement of the corticospinal tract (anterior and posterior limb of the internal

capsula).

Statistical analysis

For statistical analysis ADC, FA, and eigenvalues calculated by the software were used. Am

values were calculated using the eigenvalues A.l, A.2 and A.3 in the following way:

A-r-tCA-2 + A-3)

AI+ A-2+ A3

To determine intra-observer reliability of region-of-interest placement, the same observer

repeated the placement in a subset of 10 scans. Single measure intra class correlation coefficients

(ICC) were calculated to compare the variability of data obtained.

CH \PT!cR 6 121

To explore mean group differences, we averaged values obtained of the left and right

hemisphere per region, per subject. All regions were then divided into the following groups

based on NF1 and on the prevalence of UBOs: (1) NF1 regions with bilateral UBOs or with a

UBO on 1 site and a normal-appearing contralateral site, (2) NF1 regions with no UBOs (both

sites normal appearing); and (3) healthy controls. Differences in ADC, FA, Am, and eigenvalues

were compared by using 1-way analysis of variance (AN OVA) or the Kruskall-Wallis test (if the

distribution of the data was skewed). Significance was set at p < 0.05 and post hoc comparisons

were done by using Scheffe. Differences in the proportions of A.1 and \2, as well as the

proportion of \1 and \3, between HI hyperintensities and UBOs in other regions, were analyzed

using Wilcoxon signed ranks test for non parametric related samples. Statistical analysis was done

by using the Statistical Package for the Social Sciences 10.0 for Windows (SPSS, Chicago, Ill).

Figure 1: Region-of-interest placement.

Transverse b = 0 s/mm' images of a healtby control (multi repetition, single shot echo-planar sequence; slice thickness of

3 mm witb no gap; 25 gradient directions; b = 1000 s/mm2; TR/TE = 15000/82.1 ms; 1 average; field of view of 240 x

240 mm'; matrix of 128 x 128; voxel size of 1.8 x 1.8 x 3.0 mm3). Circular Regions-of-Interest are placed in tbe (A)

cerebellar white matter, (B) cerebral peduncle and hippocampus, (C) tbalarnus, (D) globus pallidus, and (E) frontal and

parieto-occipital white matter.

122 QUANTITATIVE DIFFERENTIATION BETWEEN HEALTHY AND DISORDERED BRAIN

MATTER IN NEUROFIBROMATOSIS TYPE 1 PATIENTS USING DIFFlJSION TENSOR

IMAGING

Figure 2: Diffusion Tensor Images of the globus pallidus.

Transverse Diffusion Tensor Images (multi repetition, single shot echo-planar sequence; slice thickness of 3 mm with no

gap; 25 gradient directions; b = 1000 s/mm'; TR/TE = 15000/82.1 ms; one average; field of view of 240 x 240 mm2;

matrix of 128 x 128; voxel size of 1.8 x 1.8 x 3.0 mm'). Girl with NF1 (age 13 years) with an unilateral UBO in the

Globus Pallidus. Arrow indicates an area of high intensity on the b = 0 mm2/s image, and high values on the ADC-, 1..e­

and ).3-maps.

Results

Intra-observer reliability of region-of-interest placement

Reproducibility of region-of-interest placement by a single observer proved moderate to very

good, with ICCs between 0.47 and 0.91. The lowest values were found for the GP (left

hemisphere, 0.47; right hemisphere, 0.70), HI (left, 0.66; right, 0.55), and TH (left, 0.52; right,

0.73), whereas the highest ICCs were found for CWM (left, 0.67; right, 0.91). However, the wide

range ofiCCs in these regions is disconcerting. In FWM, POWM and CP the ICC was >0.71.

CH.\PTFR() 123

Qualitative evaluation of UBOs

Controls (4 girls, age range: 7.4-12.1 years, mean age 10.8 years and 4 boys, age range: 7.8-12.0

years, mean age 9.6 years) were selected for comparison. All controls were without chronic

disease and had normal findings on :MR imaging.

Sixty-eight percent of the children with NF1 (n = 34) were found to have UBOs in 1 or more of

the 7 selected regions, which could be detected and measured on the b = 0 s/mm2 images.

Forty-six percent of the children with NF1 had UBOs in the GP (13 bilateral, 10 unilateral),

14% in the TH (2 bilateral, 5 unilateral), 50% in the CWM (16 bilateral, 9 unilateral) and 22% in

the CP (9 bilateral, 7 unilateral). The HI was visually scored bilaterally hyperintense on T2- and

FLAIR images in 48% of the children with NF1. No circumscript UBOs were found in POWM

and FWM. Figure 2 presents ADC, FA, and eigenvalues maps of a girl with NF1 with a UBO

in the left GP and a normal-appearing contralateral side.

Measurements in the CP could not be performed in 2 children with NF1, and in 1 of those

children, motion artefacts also prohibited measurements in the HI. Means ± SD of ADC, FA,

Am, and eigenvalues per region per group are plotted in Figures 3-6.

Quantitative differentiation of healthy and disordered brain matter in NF1

ADC values were significantly higher in regions where UBOs were present than in the matching

regions with no UBOs (for all regions, p < 0.01). Also, ADC values were higher in NF1 regions

with no UBOs than in the regions of healthy controls, significantly so in the POWM and CWM

(p < 0.03), FWM (p < 0.01), GP (p < 0.04) and TH (p < 0.01; Figure 3).

To investigate the influence of region-of-interest size, we performed additional measurements in

the GP region in a subset of 15 patients, by using a circular region-of-interest of 30 mm2• This

smaller region-of-interest size resulted in a significantly higher ADC value (p < 0.01) as

compared with the value obtained by a region-of-interest of 130 mm2•

With respect to FA values, the results were not as clear-cut. Although there was a trend toward

FA values in NF1 regions with UBOs being lower than those in matching NF1 regions without

UBOs and controls, these differences were not significant for most regions (Figure 4). Only

for CWM was the FA value in NF1 with UBOs significantly lower than that in NF1 without

UBOs (p < 0.01). Remarkably, results in the GP were opposite to those in other tissues, with

higher FA values for NF1 with UBOs as compared to NF1 without UBOs (p < 0.01).

124 QU1\NTITATIVE DIFFERENTIATION BETW:EEN HEALTHY AND DISORDERED BRAIN

MATTER IN NEUROFIBRO!'vlATOSTS TYPE 1 PATIENTS USING DIFFUSION TENSOR IMAGING

1.00

0.90

0.80

.. 0.70

l 0.80

~ 0.50 )(

0 0.40

~0.30 0.20

0.10

0.00 POWM

• UBO Cl NAS Cl control

* ,....._,

, I

CWM

* ,....._,

CP

Figure 3: Mean ADC values. The cohort is divided in tlu:ee groups: NF1 regions with UBOs (UBO), NF1 regions with

normal-appearing brain matter (NAS) and controls. Mean ADC values for UBOs are higher than those for NAS in all

regions (except POWM and FWM were no UBOs were found). Mean ADC values for NAS are significantly higher than

those of controls in all regions, except CP. For convenience, the results of the hippocampal area are presented in the same

figure as the results of the other regions-of-interest. Hippocampal hyperintensities are presented as UBO and normal

appearing hippocampal areas as NAS. Significant differences between UBO and NAS are indicated by *, between NAS

and controls by +. CP: cerebral peduncle, CWM: cerebellar white matter, HI: hippocampus, TH: thalamus, GP: globus

pallidus, FWM: frontal white matter, and POWM: parieto-occipital white matter

Am values were lower in NF1 regions with UBOs as compared with NF1 regions without

UBOs, significantly so in CWM (p < 0.02) and CP (p < 0.01). In addition, Am values were also

significantly lower in NFl regions without UBOs as compared with controls in the TH (p <

0.02). In GP, an opposite trend was shown: the value found for NFl with UBOs was

significantly higher as compared with that of NFl without UBOs (p <0.01). No significant

differences in Am were found between UBO, normal-appearing sites, and controls in the other

regions assessed. (Figure 5).

Microstructural integrity ofNFl brains

To examine microstructutal integrity of the brain parenchyma in NFl, we assessed the

eigenvalues (table 1). In all regions tested, NF1 regions with UBOs had eigenvalues higher than

those ofNF1 regions without UBOs (Figure 6). In CWM and GP all three eigenvalues

CH.\PTFR6 125

0.60 * ,......., 0.50

I

"2 .S! 0.40 ...

* c: e ,......., '6 0.30 0

rn .s. <( 0.20 u..

0.10

0.001

POWM FWM GP TH HI CWM CP

• UBO 0 NAS D control

Figure 4: FA values.

Mean FA values ± SD for UBO, normal-appearing site (NAS) and controls are plotted for the 7 regions of interest. Few

significant differences are found between UBO and NAS indicated by *. CP: cerebral peduncle, CWJ.V1: cerebellar white

matter, HI: hippocampus, TH: thalamus, GP: globus pallidus, FWM: frontal-, and POWM: parieto-occipital white matter.

0.35

0.10

0.00

POWM FWM GP TH HI GWM CP

• UBO t:11 NAS 0 control

Figure 5: Am values.

Mean Am values ± SD for UBO, normal-appearing site (NAS) and controls per regions of interest. Significant differences

between UBO and NAS are indicated by * and between NAS and controls by +. CP: cerebral peduncle, CWM:

cerebellar white matter, HI: hippocampus, TH: thalamus, GP: globus pallidus, FWM: frontal-, and POWM: parieto­

occipital white matter.

126 QUANTITATIVE DIFFERENTIATION BET\li'EEN HEALTHY AND DISORDERED BRAIN

MATTER IN NEUROFIBROMATOSIS TYPE 1 PATIENTS USING DIFFUSION TENSOR

IMAGING

were significantly higher (p < 0.02), in HI, CP and TH, Az and A3 were significantly higher (p <

0.04), indicating a loss of microstructure of the brain parenchyma in children with NF1 with

UBOs.

The eigenvalues in NF1 regions without UBOs closely followed those in healthy controls,

indicating that the microstructure is close to normal in children without UBOs, even though

some slight but significant elevations were seen (FWM, p < 0.01 for At and Az; HI, p < 0.02 for

A1). The exception is in TH, where the eigenvalues of NF1 without UBOs are much higher than

those of healthy controls, significantly so for Az and A3 (p < 0.01).

Table 1: DTI values per ROI)*

ROI Group N Mean Value ADCx10· AIX 10·3 A2X 1Q-3 A3x 10·3 3rnm.2/s FA mm2/s mm2/s mm2/s Am

POWM NAS 50 0.79 ± 0.05 0.31 ± 0.06 1.09 ± 0.06 0.74 ± 0.06 0.52 ± O.Q7 0.20 ± 0.04 Controls 8 0.75 ± 0.02 0.36 ± 0.06 1.06 ± 0.04 0.70 ± 0.05 0.48 ± 0.03 0.19 ± 0.06

FWM NAS so 0.80 ± 0.04 0.31 ± 0.04 1.04 ± 0.05 0.76 ± 0.04 0.56 ± 0.05 0.16 ± 0.02 Controls 8 0.75 ± 0.03 0.30 ± 0.02 0.98 ± 0.04 0.71 ± 0.04 0.53 ± 0.04 0.16 ± 0.02

GPt UBO 23 0.83 ± O.G7 0.25 ± 0.06 1.02 ± 0.12 0.79 ± O.G7 0.64 ± 0.06 0.13 ± 0.04 NAS 27 0.72 ± 0.04 0.17 ± 0.04 0.83 ± 0.05 0.70 ± 0.04 0.60 ± 0.04 0.09 ± 0.01 Controls 8 0.70 ± 0.02 0.16 ± 0.02 0.81 ± 0.03 0.68 ± 0.03 0.58 ± 0.03 0.23 ± 0.02

TH UBO 7 0.81 ± 0.04 0.29 ± 0.10 1.04 ± 0.05 0.75 ± 0.06 0.61 ± 0.08 0.15 ± 0.04 NAS 43 0.76 ± 0.03 0.31 ± 0.05 1.02 ± 0.05 0.70 ± 0.04 0.55 ± 0.04 0.17 ± 0.03 Controls 8 0.70 ± 0.02 0.36 ± 0.06 0.98 ± O.Q7 0.63 ± 0.03 0.47 ± 0.02 0.21 ± 0.04

HI UBO 23 0.90 ± 0.05 0.20 ± 0.04 1.07 ± 0.51 0.87 ± 0.42 0.72 ± 0.51 0.10 ± 0.03 NAS 25 0.84± 0.04 0.21 ± 0.04 1.04 ± O.Q7 0.82 ± 0.04 0.66 ± 0.04 0.12 ± 0.03 Controls 8 0.80 ± 0.02 0.21 ± 0.02 0.97 ± 0.02 0.78 ± 0.03 0.63 ± 0.03 0.11 ± 0.02

CWMt UBO 25 0.85 ± 0.14 0.32 ± 0.11 1.11 ± 0.08 0.85 ± 0.35 0.61 ± 0.13 0.20 ± 0.18 NAS 25 0.72 ± 0.06 0.43 ± 0.11 1.05 ± 0.13 0.65 ± 0.07 0.47 ± 0.09 0.23 ± O.Q7 Controls 8 0.68 ± 0.02 0.45 ± 0.08 1.00 ± 0.06 0.62 ± 0.05 0.43 ± 0.04 0.23 ± 0.05

CP UBO 16 0.89 ± 0.05 0.44 ± 0.06 1.31 ± 0.09 0.77 ± 0.09 0.54 ± 0.09 0.25 ± 0.04 NAS 33 0.81 ± 0.05 0.47 ± 0.05 1.32 ± 0.11 0.69 ± 0.09 0.47 ± 0.05 0.30 ± 0.05 Controls 8 0.80 + 0.05 0.51 ± 0.04 1.32 + 0.12 0.63 + 0.05 0.43 + 0.04 0.33 + 0.03

ROI indicates region of interest; NAS, normal-appearing sires; POWM, parieto-occipital white matter; FWM, frontal

white matter; GP, globus pallidus; TH, thalamus; HI, hippocampus; CWM, cerebellar white matter; CP, cerebral peduncle;

UBO, unidentified bright object; ADC, apparent diffusion coefficient; FA, fractional artisotropy; Am, axial anisotropy.

* Values are averaged for left and right hemispheres. For all indices, mean values ± SD are given for regions with UBOs,

regions without UBOs (NAS), and healthy controls.

t Nonparametric data.

Analysis of the difference in proportion of At and Az for hyperintensities in the HI and for

UBOs in other regions revealed there were no differences in the proportion of At-Az for HI

compared with the TH (p < 0.11) or GP (p < 0.59). The same results were found for the

(JI \I'TER () 127

differences in the proportions of A.r and A-3• This indicates that the diffusion perpendicular to the

axon for HI is not different from other grey matter areas like TH and GP. Compared with white

matter areas CP and CWM, there were significantly different proportions for A.t-A-2 (p <0.02 and

p <0.01).

~ HI

·~ 1.5 ~ 1.5 .. ...----. '$! 1 (] 'g1 .. ...----. .!:!. IJ

X

IIJ c 0.5 u iQ.S _g

i 0 i Q

At X2 13 Al A2 A3

~ F'iH ·f 1.5

CWM

~E 1.5 .. .. • ...----. ...----.

E ,_....___, • ~ , IIJ * -b t

IJ ,_....___, ...----. I 0.: IJ IJ J 0.5 lli:J - 0

A1 A2 A3 ..

),1 A2 A3

CP Gl'

~~ .!!!

~ 1.5 .. ] 1.5 ...----. .. .. M~

IIJ r-"----- * "' ...-----. .. 1 ...-----. ~ ...-----. .. IIJ IIIJ i 0.5 Ill ~ 0,5 IIIII::J 7i

0 i .. 0

" 1\1 1<2 1<3 -~ A1 1<.2 i\3 "'

• LeO • NAS 0 control

i TH • 1.5 ...-----. . ... ~ • 9 ...-----. •

IIJ ...-----.

~ 0,5 II:J !! 0 !E 111 1.2 ).3 ,

Figure 6: Eigenvalues per region.

For UBO, normal-appearing sites (NAS) and controls, mean values for the three eigenvalues are plotted for each region.

In all regions were UBOs are found, :U, A.2 and A3 are higher for UBOs than NAS and higher for NAS compared with

controls. Significant differences between UBO and NAS are indicated by * and between NAS and controls by +.

Discussion

DTI in 50 Children with NFl and 8 controls revealed significantly higher ADC values in NFl

regions with UBOs as compared to NFl regions without UBOs and in NFl regions without

128 QUANTITATIVE DJFFFRENTIATJON BETWEEN HEALTHY AND DISORDERED BRAIN

MATTER IN NEUROFIBROMATOSIS TYPE 1 PATIENTS USING DIFFUSION TENSOR

llVIAGING

UBOs as compared to controls. ADC values reflect the overall brain water content,19 thus the

differences between NF1 regions with UBOs, NF1 regions without UBOs and controls found

in this study can primarily be explained by increased water content of the brain parenchyma in

NF1, which is apparently exacerbated in NF1 regions with UBOs. These findings confirm

previous reports.6-9. 14 However, the ADC values found in this study are not as high as the ADC

values found previously. An explanation of lower ADC values may be fo~d in the shorter

effective TE and different b-value that was used in our scan protocol compared to reported

DWI protocols.20 Also DTI improves the ability to avoid partial volume effect of CSF by flne

tuning region-of-interest placement by simultaneous use ofb = 0 s/mm2 images, ADC, and FA

maps. Another important technical issue is the chosen size of the region-of-interest, since it

influences the ADC values. Larger regions of interest result in lower ADC values because the

value represents an average of more voxels, the area of the UBO and surrounding tissue, as

demonstrated in the GP by the subtest in which a region-of-interest of 30 mm2 instead of 130

mm2 was used.

DTI may facilitate a better understanding of the abnormalities seen in NF1 because evaluation

of microstructural integrity of the parenchyma can be achieved by assessing anisotropy indexes

and eigenvalues. Anisotropy can be influenced by factors such as axon packing, relative

membrane permeability to water, internal axon structure, myelination, and tissue water

content.21 ADC values in this study were higher in NF1 regions with UBOs and in NFl regions

without UBOs than in healthy controls, suggesting increased tissue water content or decreased

axon packing. However, based on FA values in our study, it was only possible to differentiate

between NF1 regions with UBOs and NF1 regions without UBOs in the CWM and GP.

Remarkably, in the latter, we found higher FA values for NF1 with UBOs as compared to NF1

without UBOs and controls, which is a counterintuitive finding.

When carefully re-examining region-of-interest placements in the GP, it became clear that it is

almost impossible to avoid partial volume effects of the posterior limb internal capsula, even

when regions-of-interest were drawn smaller (30 mm2 instead of 130 mm2). UBOs are typically

found very near or in some cases in the internal capsula. The high anisotropy of the internal

capsula affects the measured FA values for the GP. Low ICC and high variability (Fig 3) also

show the difficulties of taking measurements in the GP-region. In contrast to our results in

children with NF1, a recent published study on adult NF1 patients using DTI found

significantly lower FA values in NF1 brains than in healthy brains, indicating generalized

microstructural alterations and dysmyelination in adult patients with NFl.J4 A caveat of the

Ct-L\PTFR 6 129

study in adult patients is that it did not assess alterations of the eigenvalues, and it is, therefore,

not possible to relate changes in FA to dysmyelination. Lower FA values at UBO-sites might

also be caused by damage to the axon as shown by MR Spectroscopy.22 The study in adult

patients found severely reduced concentrations of N-acetylaspartate at UBO sites, indicating

that an increased myelin tumover was present, which could lead to subsequent axonal damage in

adult NF1 patients. We found no evidence for axonal damage in children with NF1, which

might be an explanation of why we did not find lowered FA values in our study.

We also did not find differences in the shape of the diffusion ellipsoid, when looking at Am

values, between children with NFl with or without UBOs or healthy controls in most regions­

of-interest. Although of all anisotropy indexes, Am shows the strongest trend in relative

changes,23 in this study, Am was only slightly more sensitive than FA. The homogenous

attenuated white matter structures that in CWM and CP in contrast to GP, TH and HI, where

the tissue also contains grey matter which has zero anisotropy,24 could be the reason why we did

find differences in anisotropy in CWM and CP between NFl with UBOs and NF1 without

UBOs, but not in the other regions-of-interest. If the eigenvalues At and Az and/ or A3 changed in

the same direction, as found in our study, no differences in FA and Am would be observed (but

the change in ADC would be marked).25

Higher eigenvalues in NFl regions with UBOs indicate that the microstructure of the

parenchyma is different from the parenchyma in NF1 regions without UBOs. Animal studies

have shown that an increment of axial diffusivity (Az and A3) is related to myelin deficiency,

whereas a decrease of parallel diffusivity CAt) indicates axonal disturbance.26-2S Our findings of

higher values for A2 and A3 indicate that diffusion perpendicular to the white matter structure is

higher. Because we did not find lower values for At in NF1 regions with UBOs than in NF1

regions without UBOs, we hypothesize that the observed changes of the brain tissue in NF1 are

not caused by damage to the axon, but relate to myelin deficiency. The higher value for all

eigenvalues in our study, especially for At, has not previously been reported in human or animal

studies. Accumulation of fluid hypothetically should increase the magnitude of all three

eigenvalues from their normal values.29 Our study, therefore, indicates intramyelinic edema or

vacuolar changes in the myelin. This confirms the study of DiPaolo et al,s which was the only

pathologic study performed in NFL Autopsy and histopathologic examination performed on 3

patients with NFlrevealed intramyelinic vacuolar changes or spongiotic myelinopathy that

correlated with the hyperintensities found on T2-weighted images. No stainable material was

found within the vacuoles, which suggests that in life, they are fllled with water.s

130 QUANTITATIVE DTFFERENTlt\TlON B.ETWEEN HEALTHY AND DISOllllERED BRAIN

1\IATTER IN NEUROFIBRO!vL'\TOSIS TYPE 1 PATIENTS USING DIFFUSION TENSOR IMAGING

Notably, our study is not a radiology-pathology correlation study. To our knowledge, the

specificity of changes of eigenvalues due to myelin disturbance in humans has not been tested.

Because we were not able to perform histopathological examination in our cohort, no proof can

be provided for the suggestion of myelin deficiency or vacuolar changes, however likely that

eigenvalues will allow an assessment of myelin.

HI hyperintensities were of special interest in our study, since a different pathogenetic basis

from classical UBOs was suggested.1S Impairments in learning and behavior in mouse models of

NF1 are thought to be suggestive of disordered hippocampal functioning.ls We measured DTI

parameters to examine if there is an underlying microstructural change in the hippocampal

region distinct from other regions-of-interest. We found higher ADC values in hyperintense­

appearing hippocampal areas than in normal-appearing hippocampal areas, which could explain

disordered hippocampal functioning, but no differences in FA and Am. Although anisotropy

indices in HI were lower compared to other regions-of-interest, eigenvalues showed no different

pattern when comparing HI to other grey matter structures like GP and TH. Therefore, distinct

pathogenesis between hyperintense HI and classical UBOs cannot be concluded in this study by

using DTI-parameters.

Our study has several limitations such as limited number of Children with NF1 with and

without UBOs and even smaller number of control subjects. However, post-hoc power analyses

showed that power was > 0.80 for all analyses in which NF-1 children with and without UBOs

were compared with healthy controls for ADC values. Although our findings contribute to the

unraveling of UBOs, for we have been able to prove that the high ADC values in UBOs

observed in previous publications are due to increased axial diffusivity, no histological

correlation with the observed diffusion signal abnormalities in NF1 could be provided.

Conclusion

Based on the results obtained in the current study, it can be concluded that it is possible to

differentiate quantitatively between healthy and disordered brain parenchyma in children with

NF1 using ADC values. Although no differences were found in anisotropy indexes, higher

values for Az and A3 in NF1 regions with UBOs than in NF1 regions without UBOs indicate

higher axial diffusivity because of less obstruction (presumably due to water accumulation in

myelin). The higher A1 contradicts axonal disturbance. The observed high diffusivity for all 3

eigenvalues (A1, Az and A3) in NF1 regions with UBOs, as compared to NF1 regions without

Cil \PrER 6 131

\

UBOs and\ controls, supports pathological findings, and could indicate a disturbed

microstructure of the NF1-brain due to accumulation of fluid or vacuolation when UBOs are

present. A distinct pathogenesis between hyperintense HI and classic UBOs was not found in

this study using by DTI-parameters.

Acknowledgements

We thank all children and their parents for their participation, all technicians who performed

MR imaging and the representatives of the NFl clinical workgroup in the departments of

pediatric dermatology, pediatric ophthalmology, general pediatrics, pediatric neurology and

clinical genetics.

132 QUANTITATIVE DJFFERENTJATJON BETWEEN HEALTHY AND DISORDERED BRAIN

MATTER IN NEUROFIBROMATOSIS TYPE 1 PATIENTS USING DIFFlJSION TENSOR 11\>L'\GlNG

References

1. Huson, S.M., Compston, D.A., Harper, P.S. A genetic study of von Recklinghausen neurofibromatosis in south east

Wales. I. Prevalence, fimess, mutation rate, and effect of parental transmission on severity. J Med Genet

1989;26(11):704-11.

2. Aoki, S., Barkovich, AJ., Nishimura, K, et a!. Neurofibromatosis types 1 and 2: cranial MR findings. Radiology

1989;172(2):527-34.

3. North, K .. Neurofibromatosis type 1. Am] Med Genet 2000;97:119-127.

4. Menor, F., Marti-Bonmati, L., Mulas, F., et a!. Imaging considerations of central nervous system manifestations in

paediatric patients with neurofibromatosis type 1. Pediatr Radiol1991;21(6):389-94.

5. DiPaolo, D.P., Zimmerman, R.A., Rorke, L.B., eta!. Neurofibromatosis Type I: pathologic substrate of high-signal­

intensity foci in the brain. Radiology 1995;195:721-724.

6. Mori, H., Abe, 0., Okubo, T., et a!. Diffusion property in a hamartomatous lesion of neurofibromatosis type 1. J

Comput Assist Tomogr 2001;25(4):537 -9.

7. Eastwood, J.D., Fiorella, D.J., MacFall, J.P., eta!., Greenwood, R.S. Increased brain apparent diffusion coefficient in

children with neurofibromatosis type 1. Radiology 2001 (219):354-358.

8. Alkan, A., Sigirci, A., Kudu, et a!. Neurofibromatosis type 1: diffusion weighted imaging findings of brain. Eur J

Radio! 2005;56(2):229-34.

9. Tognini, G., Ferrozzi, F., Garlaschi, G., eta!. Brain apparent diffusion coefficient evaluation in pediatric patients with

neurofibromatosis type 1. J Comput Assist Tomogr 2005;29(3):298-304.

10. Beaulieu, C. The basis of anisotropic water diffusion in the nervous system - a technical review. NMR Biomed

2002(15):435-455.

11. Le Bihan, D., Van Zijl, P. From diffusion coefficient to the diffusion tensor. NMR Biomed 2002(15):431-434.

12. Song, S., Sun, S., Ramsbottom, MJ., et a!. Dysmyelination revealed through MRI as increased radial (but unchanged

axial) diffusion of water. Neuroimage 2002(17):1429-1436.

13. Song, S., Yoshino, J., Le, T.Q., eta!. Dymyelination increases radial diffusivity in corpus callosum of mouse brain.

Neuroimage 2005(26):132-140.

14. Zamboni, S.L., Loenneker, T., Boltshauser, E., et a!. Contribution of diffusion tensor MR imaging in detecting

cerebral microstructural changes in adults with neurofibromatosis type 1. Am J Neuroradiol2007;28(4):773-6.

15. Gill, D.S., Hyman, S.l., Steinberg, A., North, K.N. Age-related findings on MRI in neurofibromatosis type 1. Pediatr

Radiol2006;36:1048-1056.

16. National Institutes of Health Consensus Development Conference Statement: Neurofibromatosis. Bethesda, Md,

USA, July 13-15, 1987. Neurofibromatosis 1988;1(3):1421-1425.

17. Basser, P.J., Mattiello, J., Le Bihan, D. Estimation of the effective self-diffusion tensor from the NMR spin echo. J

Magn Reson B 1994(103):247-254.

18. Conturo, T.E., McKinstry, R.C., Akbudak, et a!. Encoding of anisotropic diffusion with tetrahedral gradients: a

general mathematical diffusion formalism and experimental results. J Magn Reson Med 1996(32):399-412.

19. Neil J.J., Shiran S.I., McKinstry R.C., et a!. Normal brain in human newborns: apparent diffusion coefficient and

diffusion anisotropy measured by using diffusion tensor MR imaging. Radiology 1998(209):57 -66.

20. Szafer, A., Zhong, J., Anderson, A.W., Gore, J.C. Diffusion-weighted imaging in tissues: theoretical models. NMR

Biomed 1995;8(7-8):289-96.

21. Basser, P J., Jones, D.K. Diffusion-tensor MRI : theory, experimental design and data analysis - a technical review.

NMR Biomed 2002;15(7 -8):457 -67.

C:H.\PTER (i 133

22. Wang, P.Y., Kaufmann, W.E., Koth, C.W., et al. Thalamic involvement in neurofibromatosis type 1: evaluation with

proton magnetic resonance spectroscopic imaging. Ann Neurol2000;47(4):477-84.

23. Van Pul, C., Buijs, J., Janssen, M.J.A, et al. Selecting the best index for following the temporal evolution of apparent

diffusion coefficient and diffusion anisotropy after hypoxic-ischemic white matter injury in neonates. Am J

Neuroradiol 2005;26:469-481.

24. Shimony, J.S., McKinstry, R.C., Akbudak, E., et al. Quantitative diffusion-tensor anisotropy brain l'viR imaging:

nonnative human data and aoatomic aoalysis. Radiology 1999;212(3):770-84.

25. Bonekarnp, D., Nagae, L.M., Degaonkar, l'vl, et al. Diffusion tensor imaging in children and adolocents:

reproducibility, hemispheric, aod age-related differences. Neuroimage 2007;34(2):733-42.

26. Ono, J., Harada, K., Mano, T., et al. Differentiation of dys- aod demyelination using diffusional anisotropy. Pediatr

Radiol1997;16(1):63-66.

27. Takahashi, M., Ono, J., Harada, K, et al. Diffusional Anisotropy in cranial nerves with maturation: quaotitative

evaluation with diffusion MR Imaging in rats. Radiology 2000;216:881-885.

28. Nair, G., Taoahasi, Y., Low, H.P., et al. Myelination aod long diffusion times alter diffusion-tensor-imaging contrast

in myelin-deficient shiverer mice. Neurolmage 2005;28:165-174.

29. Mukherjee, P., Miller, J.H., Shimony, J.S., et al. Diffusion-tensor MR imaging of gray aod white matter development

during normal humao brain maturation. AJNR Am J Neuroradiol. 2002;23:1445-56.

134 QlTANTlTATJVE DIFFERENTIATION BETWEEN HEALTHY AND DISORDERED BRAIN

MATTER IN NEUROFIBRO!I<fATOSlS TYPE 1 PATIENTS USING DIFFlJSION TENSOR IMAGING

1-5: The NFJ CoRe Team {Cognitive Research Team), Erasmus MC University kledical CeJJter­Sophia Children's Hospit:ll, Rotterdam, The Netherlands, departments o[1General ~e,diatdcs,

2 Neuroscience, 3 Pediatdc Neurology, 4 Public Healtb and 5 Pediatdc Radiologyj 6 Erasmus MC UniFersity Medical Center, Rotterdam, Tbe NetherlaJJds, department o[ Psychiatry

7UCLA, Los Angeles, Ca!i[omia 90095-1761, USA, departments o[Neurobiology; Ps_ychiatry and Psychology, Brain Research l11stitute.

JAMA, 2008;300(3): 287-294

Abstract

Context

Neurofibromatosis type 1 (NF1) is among tbe most common genetic disorders tbat cause

le~rning disabilities. Recently, it was shown tbat statin-mediated inhibition of 3-hydroxy-3-

metbylglutaryl coenzyme A reductase restores tbe cognitive deficits in an NF1 mouse model.

Objective

To determine tbe effect of simvastatin on neuropsychological, neurophysiological, and

neuroradiological outcome measures in children witb NF1.

Design, setting and participants

Sixty-two out of 114 eligible children (54%) witb NF1 participated in a randomized, double­

blind, placebo-controlled trial conducted between January 20, 2006, and February 8, 2007, at an

NF1 referral center at a Dutch university hospital.

Intervention

Simvastatin or placebo treatment once daily for 12 weeks.

Main Outcome Measures

Primary outcomes were scores on a Rey complex figure test (delayed recall), cancellation test

(speed), prism adaptation, and tbe mean brain apparent diffusion coefficient based on magnetic

resonance imaging. Secondary outcome measures were scores on tbe cancellation test (standard

deviation), Stroop color word test, block design, object assembly, Rey complex figure test

(copy), Beery developmental test of visual-motor integration, and judgment of line orientation.

Scores were corrected for baseline performance, age, and sex.

Results

No significant differences were observed between tbe simvastatin and placebo groups on any

primary outcome measure: Rey complex figure test (B=0.10, 95% confidence interval [CI]: -0.36

to 0.56); cancellation test [speed] (B=-0.19, 95% CI: -0.67 to 0.29); prism adaptation (odds

ratio=2.0; 95% CI 0.55 to 7.37) and mean brain apparent diffusion coefficient (!3=0.06, 95% CI:

-0.07 to 0.20). In tbe secondary outcome measures, we found a significant improvement in tbe

simvastatin group in object assembly scores (B=0.54, 95% CI: 0.08 to 1.01), which was

138 THE NFl SIMVASTATIN TRIAL

specifically observed in children with poor baseline performance (13=0.80, 95% CI: 0.29 to 1.30).

Other secondary outcome measures revealed no significant effect of simvastatin treatment.

Conclusions

In this 12-week trial, simvastatin did not improve cognitive function in children with NFl.

Trial registration

Trial registration isrctn.org Identifier: ISRCTN14965707

Introduction

Neurofibromatosis type 1 (NF1) is a common autosomal-dominant genetic disorder (incidence

1:3000)1 caused by a mutation in the gene encoding neurofibromin, a protein that activates the

hydrolysis of RAS-bound guanosine triphosphate.z NF1 is characterized by various

neurocutaneous manifestations, problems in fine and gross motor functioning3, as well as the

frequent occurrence of cognitive disabilities. Children with NF1 have a lowered mean IQ (86-

94), with particular deficits in visual-spatial skills, nonverbal long-term memory, executive

functions and attention.4-7 These problems have a large impact on school performance of

children with NF1.4 It has been suggested that the cognitive and motor deficits in children with

NF1 are related to hyperintensities on T2-weighed magnetic resonance imaging of the brain3, 8

that are characterized by high apparent diffusion coefficients (ADC values),9 but some studies

fail to confirm this relationship.ID

Studies using mouse models for NF1 (Nf1 mice) revealed that increased RAS/ERK. signaling is

primarily responsible for the neuronal plasticity deficits as well as the spatial learning and

attention deficits of these mice.l1-13 RAS transforming activity requires isoprenylation (i.e.,

farnesylation or geranylgeranylation) of RAS, which can be blocked by farnesyl transferase

inhibitors and by 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG-CoA) reductase

inhibitors.14. 15 HMG-CoA reductase is the rate-limiting enzyme in the mevalonate pathway in

which cholesterol and isoprenyl groups are synthesized. Importantly, treatment of Nf1 mice

with a farnesyl transferase inhibitor or HMG-CoA reductase inhibitor for just a few days,

reverses the cognitive deficits of these mice. II, 13 These findings are not only important for NF1,

but also are of great interest for other neuro-cardio-facial-cutaneous syndromes (e.g. Noonan,

Costello and cardio-facial-cutaneous syndromes), which are also caused by aberrant RAS/ERI<

C:H,\PTFR 7 139

signaling, and for hamartoma syndromes (e.g., Cowden disease and tuberous sclerosis complex).

The genes associated with these syndromes belong to a pathway that is not only coregulated by

RAS but also critically dependent on RHEB, another farnesylated protein of the RAS family.

The favorable safety profile of the HMG-CoA reductase inhibitor simvastatin in adults and

children16 provided an opportunity to investigate whether the findings in the mouse model can

be translated to humans. In a randomized, double-blind, placebo-controlled trial, we studied the

effect of a 12-week simvastatin treatment on cognitive function of children with NF1 using

neuropsychological, neurophysiological, and neuroradiological outcome measures.

Methods

Design

A prospective double-blind, placebo-controlled, randomized, single-site, 12-week clinical trial

was conducted in children with NF1 between January 20,2006, and February 8, 2007. The study

was approved by the medical ethical committee of the Erasmus MC Rotterdam, the

Netherlands.

Participants

All participants were recruited from the multidisciplinary NF1 outpatient clinic of the Erasmus

MC - Sophia Children's Hospital, which is a university hospital and NF1 referral center in the

Netherlands. Participants were enrolled by a pediatrician in the NF1 outpatient clinic (A.G.B.).

Inclusion criteria were age 8 to 16 years, NF1 diagnosis according to the criteria of the National

Institutes of Health,17 and oral and written informed consent from parents and children older

than 12 years. Exclusion criteria were segmental NF1, pathology of the central nervous system

(other than asymptomatic gliomas), deafness, severely impaired vision, use of antiepileptic drugs,

insufficient comprehension or use of the Dutch language, and an IQ below 48, which was

assessed at baseline using the Wechsler Intelligence Scale for Children - Revised, Dutch

version. IS

Protocol

Patients were randomized to simvastatin or placebo using a permuted-block, 1:1 randomization

list generated by the trial statistician (S.M.F.P.) with blocks of 6 participants, in which

medication numbers 1 through 62 corresponded to either simvastatin or placebo.

Randomization was performed by the Erasmus MC trial pharmacist who assigned patients a

140 THE NFl SIMVASTATIN TRIAL

medication number in the order of their enrollment in the ttial and who dispensed the

medication. Patients and all other investigators were blind to the treatment allocation. Patients

were treated once a day in the morning for 12 weeks with sirnvastatin (weeks 0-4, 10 mg/ d,

weeks S-8 20 mg/d, and weeks 9-12,20 mg/d for children aged 8-12 years or 40 mg/d [taken as

2 20-mg doses] for children aged 13-16 years) or equivalent placebo. The placebo capsules were

filled with microcrystalline cellulose PH102 and treatment capsules with a filler and a tablet of

10-mg (weeks 0-4) or 20-mg (weeks 5-12) sirnvastatin (filln-coated; Alpharma Inc; Bridgewater,

New Jersey). The capsules containing placebo or sirnvastatin were non-transparent and identical

in color, shape, and size. Patients were instructed not to open the capsules. Patients were judged

adherent when they took at least 80% of their study medication during the intervention period

of 12 weeks, which was assessed by counting returned capsules.

Outcome measures

Outcome measures were assessed at baseline and after 12 weeks of treatment. For the primary

outcome measures, we chose 2 neuropsychological tests that were analogous to statin­

responsive tests in Njt mice (measuring visual spatial memory and attention). In addition, we

selected a neurophysiological and neuroradiological measure because we reasoned that these

measurements would be insensitive to placebo or test-retest effects. This resulted in the

following 4 primary outcome measures: performance on the Rey complex figure test (CPT)

(delayed recall; assessing nonverbal long-term memory); performance on the cancellation test

(speed, assessing attention), performance on a prism adaptation task (measurement of

adaptation of the angle of hand movements in response to prism glass distortion,19 which is

thought to be dependent on cerebellar function2°. 21), and mean apparent diffusion coefficient

(ADC value) of the brain (mean ADC value of 7 predetermined anatomic locations

predominantly affected by T2-weighed hyperintensities) as previously described.9

For the secondary outcome measures, we selected neuropsychological tests assessing domains

that are specifically affected in patients with NF1: tests for attention and tests for visual-spatial

skills with baseline scores of 1 SD or more below average.4, 9 This resulted in the following

secondary outcome measures: performance on the cancellation test (standard deviation;

measuring attention fluctuations), the Stroop color word test, the block design test and object

assembly test of the Wechsler Intelligence Scale for Children- Revised, the Rey CPT (copy), the

Beery developmental test of visual-motor integration, and the judgment of line orientation

task.22

141

Magnetic Resonance imaging was performed by usmg a 1.5-tesla system (EchoSpeed; GE

Healthcare, Milwaukee, Wisconsin) and a dedicated 8-channel head coil. Diffusion tensor

imaging data were gathered by using a multirepetition, singleshot echo-planar sequence with a

section thickness of 3 mm with no gap. A 25-gradient directions technique was performed to

obtain good diffusion tensor images (sensitivity, b=1000 s/mm2, repetition time 15000 ms; echo

time, 82.1 milliseconds, 1 average; field of view, 240x240 mm2; matrix 128x128; voxel size,

1.8x1.8x3.0 mm3) as described previously.9

All neuropsychological tests were developed for children, administered in their Dutch versions,

and scored by 1 pediatric neuropsychologist (M.J.B.). Parallel versions of tests were applied

when available to reduce the impact of practice effects. For technical reasons, left-handed

children (n=7) were excluded from the prism adaptation task.

Treatment safety and adherence was assessed in the outpatient clinic at baseline, and after 4 and

12 weeks, and with a telephone consult after 8 weeks. Patients were provided with a diary in

which they were instructed to note any deviations from treatment protocol and possible adverse

events. At each consult, a general pediatrician recorded any adverse events and serious adverse

events (adverse events that were life-threatening, causing disability, or requiring hospitalization)

with a standardized checklist of the adverse events and serious adverse events reported with

simvastatin use,23 supported by open questions and a review of the patient's diary. All reported

adverse events were scored as being not drug related, possibly drug related, or definitely drug

related prior to unblinding. During the visits to the outpatient clinic, the pediatrician (A.G.B.)

performed a standardized internal and neurological assessment, and blood was drawn for

laboratory examination. We examined safety parameters alanine aminotransferase, aspartate

aminotransferase and creatine phosphokinase, and efficacy parameters total cholesterol, high­

density lipoprotein cholesterol, low-density lipoprotein cholesterol and triglycerides were

examined according to standard clinical laboratory protocol. Criteria for discontinuation of

study medication were a persistent of more than 3-fold the upper limit of normal (ULN)

increases in alanine aminotransferase or aspartate aminotransferase levels, more than 10-fold the

ULN for creatine phosphokinase with or without muscular symptoms, or 5- to 10-fold the ULN

for creatine phosphokinase levels with muscular symptoms.16

Statistical analyses

One of the prominent effects seen in statin-treated Njl mice was a recovery of their deficit in

visual-spatial memory.13 The Rey CFT (delayed recall) assesses the analogous domain of

142 THE NFl SIIYIVASTATIN TRLU.

nonverbal long-term memory in humans and has good psychometric properties, and

performance on this test is specifically affected in patients with NF1.24 Therefore, we based our

power calculation on this test. On the assumption of a correlation of 0. 70 between measurement

before and after treatment, and a mean (SD) z-score of -1.32 (1.01) on the Rey CPT (delayed

recall) at baseline,24 we calculated that 30 persons were needed in both the placebo and

treatment groups to ensure a power of 0.80 of detecting a significant (o:=0.05) improvement in

the Rey CFT (delayed recall) score up to -0.28 (difference of 1.04) in the treatment group.

All data were analyzed using SPSS 12.0 (SPSS Inc, Chicago, Illinois). For the neuropsychological

tests, z-scores were used (with negative values indicating performance below the normative

mean and positive values performance above the normative mean), except for the cancellation

test (standard deviation) (raw score for non-normal distribution of reference values; larger

negative values indicated larger attention fluctuations). Prism adaptation was scored to occur if

the change (adaptation) of the angle of hand movements was significant (p<0.01) and larger

than -1 SD of the mean change of age-matched healthy controls (n=16, unpublished

observations). A decrease in ADC values indicates lower signal intensity.

Modified intention-to-treat analysis was performed for all patients with available 12-week test

scores (n=61) without imputing missing values. Differences between the simvastatin and

placebo groups at baseline were analyzed with the t-test, Mann-Whitney test, and y} test.

Differences between the simvastatin and placebo groups after 12 weeks of treatment were

assessed using univariate and multivariate regression analysis. In the univariate analysis, we

adjusted for baseline scores, and in the multivariate regression analysis we adjusted for baseline

scores, age, and sex. Regression coefficients (13) reflect the estimated differences in mean score

at follow-up between the treatment groups with 95% confidence intervals (Cis). For categorical

measures (prism adaptation), the difference between the treatment groups was expressed as an

odds ratio with 95% CI. Cut-offlevel for significance was set at p<0.05. Effect modification of

outcome parameters that were significantly different between the treatment and placebo groups

after 12 weeks was examined using interaction terms between treatment and age and between

treatment and baseline performance. The rationale for this were the conceivably higher brain

plasticity in younger children and more room for improvement in children with low baseline

performance, thus affecting the magnitude of response to simvastatin treatment. Subgroup

analysis was performed only if effect modification was plausible (p<0.10 to take into account

the small size of the subgroups) for addition of the interaction term to the multivariate analysis.

All p-values reported are 2-sided. The outcome parameters and the method of statistical

Ul\PTER 7 143

analysis, including the subgroup analyses, were defmed before unblinding. We did not correct

for multiple comparisons for the following reasons. There are only 4 primary outcome

measures, and they are specifically based on a priori assumptions. The outcome measures on the

neuropsychological tests are potentially correlated, and correction would thus be inappropriate.

By correcting for multiple comparisons, it would be very hard to detect a possible effect in a

relatively small patient population. Thus we would run a high risk of discarding a promising

drug while in fact there is an effect (Type II error).

For ethical reasons, an interim analysis was conducted by the statistician (S.M.F.P.) after 36

patients completed the study with complete maintenance of the double-blind protocol for all

others. The criterion to discontinue the study was a significant difference between the

simvastatin- and placebo groups on Rey CFT (delayed recall) score at 12 weeks (p<0.01). The

statistician communicated that this criterion was "not reached" and the study was continued as

planned.

12 ExcludEd 3 ~rrtaJ neuro!tromatosls We 1

,___ 3 Used an~<;plep~cs 3 H)ldrocel*lokls 1 ~-mEnal rWII'dallon 2 lne«ndent use or the Dtrtch language

I 114 81glbl& J 52 Decln«< to partlclpale

13 Trial loo tlme-OOilSUTJIIlg 1or a"!Jid 6 Parents afraid 6 Oilc!a!rad - 5 Parents del oot lee! chid had ~problems

' 10 Oltler 12 No reason spECiff«<

Figure 1: Flow chart of patient inclusion.

144 THE NFl SIMVASTATIN TRL.I.L

Results

Participants

One hundred fourteen children were eligible for this study. Consent to participate was obtained

for 62 children (response rate, 54%). The children who participated in the trial (n=62) did not

differ significantly from the total eligible group (n=114) on age, sex, frequency of mental

retardation, or disease severity (all p>0.3) indicating that they were representative of the total

eligible group. The 62 participants were randomly assigned to the simvastatin group (n=31) or

the placebo group (n=31) (see Figute 1). The baseline characteristics were similar between the

simvastatin- and placebo groups for all baseline parameters except for median age (see tables 1

and 2). Mean (SD) treatment duration was 12 weeks and 3 days (6 days). There were no

deviations from random allocation. One participant (2%) in the simvastatin group withdrew

from the study after 10 weeks for personal reasons. Three of 62 children (5%), all in the placebo

group, were not adherent according to the SO% criterion. We could not retrieve all of the

medication jars for 10 of 62 children (16%, 6 in the simvastatin group and 4 in the placebo

group).

Table 1. Baseline chru:acteristics of the study groups•

Placebo (n-31) n%

Simvastatin (n-31) n%

Patient chatacteristics Age at randomization in yeru:s, median (IQR) b

Male sex Full Scale IQ, mean (SD) NF1 disease severity<

:Minimal Mild Moderate Severe

Inheritance of NF1 Familial Sporadic Unconfinned

Socio-economic statusd Low Middle High

Total cholesterol, mean (SD), mg/dL LDL cholesterol, mean (SD), mg/ dL Treatment dose week 9-12c

11.5 (9.4-13.5) 16 (52) 85 (15)

10 (32) 13 (42) 8 (26)

14(45) 16 (52) 1 (3)

12 (39) 9 (29)

10 (32) 166 (31) 97 (26)

Wmg/~ ~ ~mg/~ ~

Ma.'<imal treatment dose in rng/kg, mean (SD) NA NF1: Neurofibromatosis type 1; LDL: Low-density lipoprotein-Cholesterol; NA: not applicable. SI conversion factors: to convert cholesterol to mmol/L, multiply by 0.0259 0N=62 unless otherwise indicated. bp=0.03 between simvastatin and placebo group

13.2 (11.3-15.2) 19 (61) 88 (15)

12 (39) 11 (35) 7 (23) 1 (3)

12 (39) 19 (61) 0 (0)

12 (39) 9 (29) 10 (32) 163 (36) 96 (32)

12 (39) 19 (61)

0.7 (0.1)

cDisease severity ofNF1 was scored according to the Riccatdi scale modi£ed to exclude cognitive aspects ofNF1.4 dSocioeconomic status was detertnined from highest parental occupation or, if not available, highest parental education, and divided into low, middle, or high. '

CHAPTER 7 145

Table 2: Scores on Primary and Secondary Outcome Measrues at Baseline and 12 Weeks Baseline• 12 weeksb Univariate Multivariate

(mean (mean ___ di.fii:::':::::e7re:-;nc..c;.,:e;,b-,c-----di.fii:::'""'e:::r.:cenc:c:::e=::b:-,d--(SD)) (SD)) ~ (95% CI) @ (95% Cl)

Primary Outcome Measures Rey CFT - delayed recall

Placebo Simvastarin

Cancellation test- speed• Placebo Simvastatin

Significant Prism adaptation (N (%))' Placebo Simvastatin

AverageADC-value (x10-3mm2/s)h Placebo Simvastatin

Secondary Outcome Measures Cancellation test- standard deviationi

Placebo Simvastatin

Stroop - speedi Placebo Simvastatin

Block design Placebo Simvas tatin

Object Assembly Placebo Simvastatin

Rey CFT- copy Placebo Simvastatin

BeeryVlvll Placebo Simvastatin

Judgment of line orientation test

-1.6 (0.7) -1.5 (1.0) -1.7 (0.8) -1.4 (0.8)

-0.8 (1.6) 0.4 (1.1) -1.2 (1.8) -0.1 (1.4)

12 (44) 10 (37) 11 (SO) 12 (48)

8.03 (0.52) 7.97 (0.50) 8.02 (0.44) 7.91 (0.46)

-2.7 (1.2) -1.9 (0.9) -2.8 (1.7) -2.0 (1.5)

-0.2 (1.8) 0.2 (1.5) -0.5 (2.1) 0.3 (1.9)

·1.1 (0.8) -1.0 (1.0) -0.8 (0.9) -0.5 (1.0)

-1.1 (1.1) -0.9 (1.3) -0.8 (1.1) -0.1 (1.0

-1.2 (1.2) -0.7 (1.1) -1.4 (1.3) -1.0 (1.2)

-1.2 (0.9) -1.1 (0.9) -1.2 (0.7) -1.1 (0.7)

Placebo -1.6 (1.4) -1.1 (1.6) Simvastatin -1.1 (1.4) -0.8 (1.6)

0.07 (-0.37 to 0.51) 0.10 (-0.36 to 0.56)

-0.27 (-0.74 to 0.20) -0.19 (-0.67 to 0.29)

1.57 (0.48 to 5.13)g 2.01 (0.55 to 7.37)g

0.01 (-0.12 to 0.14) 0.06 (-0.07 to 0.20)

-0.12 (-0.65 to 0.41) -0.26 (-0.80 to 0.28)

0.34 (-0.36 to 1.04) 0.48 (-0.23 to 1.18)

0.15 (-0.18 to 0.47) 0.10 (-0.24 to 0.44)

0.50 (0.05 to 0.95)k 0.54 (0.08 to 1.01)l

-0.26 (-0.71 to 0.19) -0.12 (-0.58 to 0.34)

·0.01 (-0.27 to 0.26) -0.02 (-0.30 to 0.26)

-0.12 (-0.62 to 0.38) -0.06 ( -0.58 to 0.46)

Abbreviations: NF1: Neruofibromatosis type 1; CFT: Complex Figure Test; Beery VMI: Beery Developmental test of Visual-Motor Integration; ADC: Apparent Diffusion Coefficient. •N =62 (31 placebo, 31 simvastatin) unless otherwise indicated. Values indicate mean (SD) z-score, unless otherwise indicated, in which negative values indicate performance below the normative mean, and positive values performance above the normative mean. bN=61 (31 placebo, 30 simvastatin) unless otherwise indicated; 1 loss to follow up in the simvastatin group before final assessment. Values indicate mean (SD) z-score, unless otherwise indicated, in which negative values indicate performance below the normative mean, and positive values performance above the normative mean. •Values (regression coefficients with 95% confidence intervals) indicate between group differences in scores after 12 weeks, adjusted for baseline scores, obtained from unlvariate regression analysis. dValues (regression coefficients with 95% confidence intervals) indicate between group differences in scores after 12 weeks, adjusted for baseline scores, age, and sex, obtained from multivariate regression analysis. •Baseline and 12 Weeks: N=29 in the placebo group; only administered if children possessed sufficient rote memory to count groups of up to five dots. EJ3aseline: N=49 (27 placebo, 22 simvastatin); 7 left-handed children excluded, 6 children excluded due to technical problems including not understanding or adhering to task instructions (N=4). 12 Weeks: N=52 (27 placebo, 25 simvastatin); 6 left-handed children excluded, 3 children excluded due to technical problems including not understanding/adhering to task instructions (N=2). g0dds ratio with 95% confidence interval. N=46 (26 placebo, 20 simvastatin), 6 left-handed children excluded, 9 children excluded due to technical problems including not adhering to task instructions (N=6). hBaseline: N=SO (25 placebo, 25 simvastatin); 2 missing due to artifacts, 10 were not scanned due to limited MRI capacity (random). 12 Weeks: N=46 (23 placebo, 23 simvastatin); 5 missing due to artifacts, 10 were not scanned due to limited l\IIRI capacity (random). A decrease in ADC values indicates lower signal intensity. lRaw score. Baseline and 12 Weeks: N=29 in the placebo group; only administered if children possessed sufficient rote memory to count groups of up to five dots. Larger negative values indicate larger attention fluctuations. IBaseline: N=59 (29 placebo, 30 simvastatin); 12 Weeks: N=SS (29 placebo, 29 simvastatin), only administered if children could read the names of colors. 'p=0.03.lp=0.02.

146 THE NF1 SIMVASTATIN TRIAL

Effect of simvastatin on outcome parameters

After 12 weeks of treatment, we did not observe a significant difference between the simvastatin

and placebo groups on the primary outcome measures (Rey CFT [delayed recall], cancellation

test [speed], prism adaptation, and mean brain ADC values). We also did not observe an effect

on the secondary outcome measures (cancellation test [standard deviation], Stroop color word

test, block design, Rey CFT [copy], Beery developmental test of visual-motor integration and

judgment of line orientation), except for a higher score on the object assembly test in the

simvastatin group using univariate analysis (adjustment for baseline scores, J3=0.50 [95% CI,

0.05 to 0.95]), as well as multivariate analysis (adjustment for baseline scores, age, and sex,

J3=0.54 [95% CI, 0.08 to 1.01]) (table 2).

Paired t-tests revealed that performance after 12 weeks was similar or better than baseline for all

tests in both the simvastatin and the placebo groups. In the placebo group, the improvement

between baseline and 12 weeks was significant on 4 of 9 neuropsychological outcome measures

(cancellation test [speed and standard deviation], Rey CFT [copy], judgment of line orientation),

leading to a performance within the normal range on the first 3 tests.

Effect modification

We found that baseline performance on object assembly is a modifier of the effect of

simvastatin on this test (p=0.07). Subsequent subgroup analysis showed a significant effect of

simvastatin in the group with baseline object assembly test scores :S -1SD, (J3 =0.80 [95% CI,

0.29 to 1.30]; n=37), but not in the group with baseline object assembly score of >-1 SD (J3

=0.47 [95% CI, -0.64 to 1.59]; n=24) indicating that the difference in the object assembly test

results between the simvastatin and placebo groups is mostly caused by an increase in score in

children with a poor baseline performance in the simvastatin group (figure 2). There was no

interaction between improvement on the object assembly task and age.

Safety and effect on cholesterol levels

There were no laboratory adverse events, and no serious adverse events. In total, 5 adverse

events were reported by 3 of 31 (10%) children in the simvastatin group: hair loss (1 child after

4, 8 and 12 weeks), muscle weakness (1 child after 8 weeks), and constipation (1 child after 12

weeks) compared with 4 adverse events reported by 3 of 31 (10%) children in the placebo group

(dizziness (1 child after 4 and 8 weeks) and constipation (1 child after 8 and 1 child after 12

weeks). None of the reported adverse events reported were judged clinically significant.

CH \PTER 7 147

,...,. ~

00

>-l :I: m z "'1 >-'

"' ;: ~

~ i::i z >-1 :00

~

Low Baseline Score High Baseline Score

2.5

''l • 00

1.5

• • 1.0 • 00 00 •• • •••

•• 00 • •• T 0.5-l T • 0 •• •• • • 1 0

I ••••••

~ 0 ••• 00 0 •• •• ~ -0.5

N •• 1 00 • • -1.0 00000000

o:Oool ••• •• •

I 0000

T •••••• • •• -1.5-l 00 00 1 • 0 •

-2.0 000 00 ••• • 00 00 • 0

-2.5 0 •

-3.0 00 00 • -3.5 Baseline 12Weeks Baselne 12Weeks Baseline 12Weeks Baseline 12Weeke

I

Placebo Simvastatin Placebo Sinvastatin

Figure 2: Interaction between baseline score and effect of simvastatin on object assembly test results.

For each subgroup, individual Z scores and uncorrected group mean Z scores are provided. For each subgroup, the left range shows scores at baseline and the right range, scores at 12 weeks. For the

simvastatin group, n=16 for the low baseline score at baseline but n=15 for the low baseline score at 12 weeks; n:::::lS for the high baseline score.

For the placebo group, n:=22 for the low baseline score, and n=9 for the high baseline score. The difference between the simvastatin and placebo groups after 12 weeks is significant in the groups with

low baseline performance (!3=0.80; 95% confidence interval, 0.29 to 1.30; p=0.003), but not in the groups with high baseline performance (~=0.47; 95% confidence inten•al, -0.64 to 1.59).

Error bars represent 95°/o confidence intenT~ls.

After 12 weeks of simvastatin treatment, total cholesterol levels were reduced by a mean (SD) of

21.1% (1 0. 7%) of baseline values, and low-density lipoprotein cholesterol with 39.4% (15.1 %).

There was no significant change in high-density lipoprotein cholesterol or triglycerides. The

change in low-density lipoprotein cholesterol in the simvastatin group was not significantly

related to the dose in mg/kg, sex, or age. The low-density lipoprotein cholesterol level of the

children in the simvastatin group who did not return all of their medication jars was decreased

by at least 34% (1 not determined because ofloss to follow-up).

Comment

We report the results of a randomized, double-blind, placebo-controlled trial to investigate the

effect of simvastatin on cognitive functions in children with NFL We used a carefully selected

set of outcomes, including tests resembling measurements shown to be responsive to statins in

preclinical studies, tests reflecting the specific neuropsychological deficits in NF1, and objective

outcomes such as prism adaptation and brain ADC values, which are insensitive to a placebo or

test-retest effect. We did not find an effect of 12 weeks of simvastatin treatment on the primary

and secondary outcome parameters, except for higher scores on the object assembly test.

We can.conclude post hoc that the power of our study was enough to reject a possible effect on

most test. For instance, for the Rey CFT (13=0.10, se=0.23), we can reject a change larger than

0.56, and for the cancellation test (speed) (fi=-0.19, se=0.24), we can reject a change larger than

0.28. Furthermore, we chose to interpret an improvement of 1 SD as clinically significant, and

none of the outcome measures showed a difference between the simvastatin and placebo group

of 1 SD or larger. Thus, given the power of the study and the overall negative findings, this

study does not provide support for prescribing simvastatin to treat the cognitive deficits of

children with NF1.

The object assembly test was the only outcome measure that was significantly improved.

Considering that we only found an improvement in object assembly and that we did multiple

statistical comparisons without adjusting the p value, this is most likely a spurious finding. It

should be noted that the improvement in object assembly was restricted to children who

performed poorly at baseline. This specific improvement in the subgroup of children with poor

baseline scores is not likely to be related to a practice effect, because children with high baseline

scores are expected to benefit most from a practice effect.2s

CI-L\PTER 7 149

The object assembly test measures multiple cognitive domains, but in the context of the entire

neuropsychological assessment, along with the clinical behavioral observations made during the

assessment, visual synthesis is probably the most damaged cognitive domain. Improved visual

synthesis would affect academic performance. For instance, visual synthesis needs to be

mastered before children to start reading and spelling, and visual synthesis is an important part

of more advanced mathematics.26.27 However, whether the observed improvement in object

assembly is a real effect and whether simvastarin would indeed improve academic achievement

remain to be confirmed.

Our study has several limitations. First, the treatment duration used in our study might have

been too short to observe a clinically significant recovery in patients with NFL We based the

length of our trial on the observation that Starin treatment normalized the plasticity impairment

and cognitive phenotype of Njt mice within days,13 and the observation that treatment of

cognitive problems in children can be reached within days to weeks (for instance in the

treatment of attention deficits in attention-deficit/hyperactivity disorder, reviewed by Brown et

al2B). However, because precedents for translational trials into cognition are rare, we can not

exclude the possibility that the effect of simvastarin on higher cognitive functions in humans

would require a longer treatment period than 12 weeks.

Second, the placebo group showed a significant improvement between baseline and 12-week

scores on 4 out of 9 neuropsychological outcome measures. This resulted in a performance

within the normal range on 3 tests. Because preclinical studies showed that starin treatment did

not improve cognitive function in mice that already learned well,B it is possible that we reached

a performance ceiling, that hampered detection of an effect.

Third, it is conceivable that the therapeutic effect of simvastarin on human brain function was

hampered by suboptimal availability due to a first pass effect, or due to inefficient crossing of

the blood brain barrier. However, increasing the therapeutic dose does not seem desirable due

to the lack of safety studies in children with higher doses, and the increasing risk of side effects

observed in adults.23 Furthermore, the effect of simvastarin on low-density lipoprotein

cholesterol at 12 weeks was similar to the decrease achieved after 48 weeks of simvastarin

treatment in a previous pediatric study.16 This indicates that, at least in the liver, the treatment

dose was optimal with respect to inhibition of the mevalonate pathway.

150 THE NFl STMVASTA.TTN TRIAL

Finally, there was a relatively high number of missing data in the neuroradiological and prism

adaptation results. Although this reduces the power on these outcome measures there was no

indication for a substantial bias because the distribution of observations that were missing did

not significantly differ between the simvastatin and placebo groups. For the other outcome

measures, the proportion of missing data was negligibly small.

The overall negative outcome of this trial suggests that simvastatin should not be prescribed to

ameliorate the cognitive deficits associated with NF1. Further studies to evaluate a longer

treatment period and whether the object assembly finding is spurious may be warranted.

Acknowledgements

Author Contributions: Ms Krab had full access to all of the data in the study and takes

responsibility for the integrity of the data and the accuracy of the data analysis. Study concept

and design: I<rab, De Goede-Bolder, Aarsen, Pluijm, Lequin, Catsman-Berrevoets, Arts,

Kushner, Silva, De Zeeuw, Moll, Elgersma. Acquisition of data: I<rab, De Goede-Bolder,

Bouman, Lequin, Moll. Analysis and interpretation of data: I<rab, De Goede- Bolder, Aarsen,

Pluijm, Van der Geest, Kushner, Moll, Elgersma. Drafting of the manuscript: I<rab, De

Goede-Bolder, Van der Geest, Moll, Elgersma. Critical revision of the manuscript for

important intellectual content: I<rab, De Goede-Bolder, Aarsen, Pluijm, Bouman, Lequin,

Catsman-Berrevoets, Arts, Kushner, Silva, De Zeeuw, Elgersma. Statistical analysis: I<rab,

Pluijm, Van der Geest. Obtained funding: I<rab, De Goede-Bolder, De Zeeuw, Moll,

Elgersma. Administrative, technical, or material support: I<rab, De Goede-Bolder, Bouman,

Van der Geest. Study supervision: De Goede-Bolder, Aarsen, Lequin, Catsman-Berrevoets,

Arts, Silva, De Zeeuw, Moll, Elgersma. Financial Disclosures: Drs Kushner and Silva reported

being co-applicants on a US patent (No. 11/569,426) "Treating learning deficits with inhibitors

of HMG-CoA reductase." None of the other authors reported a potential, real, or perceived

conflict of interest or financial disclosure. Funding/Support: This trial was funded by the

Hersenstichting Nederland, the Sophia Foundation for Medical Research, and the Prinses

Beatrix Fonds and by a donation from the Dutch Neurofibromatosis Foundation. Role of the

Sponsor: The funding sources had no role in the design and conduct of the study; in the

collection, analysis, and interpretation of the data; or in the preparation, review, or approval of

the manuscript.

CH \PTER ( 151

Additional Contributions: We thank the patients and their parents for their participation.

Also, we thank E. Steyerberg, l'vlD, PhD, Department of Public Health, for fruitful discussions,

and J. B. C. De Klerk, l'vlD, Department of General Pediatrics and Clinical Genetics, for

laboratory safety monitoring. S. J. P. M. Van Engelen, Department of Pediatric Radiology; S. M.

Goorden, MSC, M. Elgersma, BSC, P. Plak, BSC, Department of Neuroscience; and A. C.

Gaemers, MA, R. Wierenga, MA, E. Barendse, MA, Department of Pediatric Neurology, helped

in collecting the data. We appreciate the support of all other participants of the NF1 clinical

workgroup and the NF1 Core (Cognitive Research) Team of the Erasmus MC University

Medical Center, Sophia Children's Hospital Rotterdam, the Netherlands. None of the

acknowledged participants received any form of compensation for their contributions.

152 THE NFl Sli'4VASTATIN TRL<\L

References

1. Huson SM, Compston DA, Clark P, Harper PS. A genetic study of von Recklinghausen neurofibromatosis in south

east Wales. I. Prevalence, fitness, mutation rate, and effect of parental transmission on severity. ] Med Genet. Nov

1989;26(11):704-711.

2. Gutmann DH, Wood DL, Collins FS. Identification of the neurofibromatosis type 1 gene product. Proc Nat!Acad Sci

US A. Nov 1 1991;88(21):9658-9662.

3. Feldmann R, Denecke J, Grenzebach M, Schuierer G, Weglage]. Neurofibromatosis type 1: motor and cognitive

function and T2-weighted MRI hyperintensities. Neurology. Dec 23 2003;61 (12):1725-1728.

4. Krab LC, Aarsen FK, de Goede-Bolder A, et al. Impact of Neurofibromatosis type 1 on School Performance. J

Child Neurol. 2008;23(9). doi:10.1177 /0883073808316366.

5. Levine TM, Materek A, Abel J, O'Donnell M, Cutting LE. Cognitive pro@e of neurofibromatosis type 1. Semin

Pediatr Neurol. Mar 2006;13(1):8-20.

6. Billingsley RL, Jackson EF, Slopis JM, Swank PR, Mahankali S, Moore BD. Functional MRI of visual-spatial

processing in neurofibromatosis, type I. Neuropsychologia. 2004;42(3):395-404.

7. North K, Hyman S, Barton B. Cognitive deficits in neurofibromatosis 1. J Child Neurol. Aug 2002;17(8):605-612;

discussion 627-609, 646-651.

8. North K,Joy P, Yuille D, et al. Specific learning disability in children with neurofibromatosis type 1: significance of

MRI abnormalities. Neurology. May 1994;44(5):878-883.

9. van Engelen SJ, I<rab LC, Moll HA, et al. Quantitative differentiation between healthy and disordered brain matter in

patients with neurofibromatosis type I using diffusion tensor imaging. A]NRAm J Neuroradiol. Apr 2008;29(4):816-

822.

10. Legius E, Descheemaeker MJ, Steyaert J, et al. Neurofibromatosis type 1 in childhood: correlation of MRI findings

with intelligence. J Neural Neurosurg Psychiatry. Dec 1995;59(6):638-640.

11. Costa RM, Federov NB, Kogan JH, et al. Mechanism for the learning deficits in a mouse model of

neurofibromatosis type 1. Nature. Jan 31 2002;415(6871):526-530.

12. Guilding C, McNair K, Stone TW, Morris BJ. Restored plasticity in a mouse model of neurofibromatosis type 1 via

inhibition of hyperactive ERIC and CREB. Eur J Neurosci. Jan 2007;25(1):99-105.

13. Li W, Cui Y, Kushner SA, et al. The HMG-CoA reductase inhibitor lovastatin reverses the learning and attention

deficits in a mouse model of neurofibromatosis type 1. Curr Bioi. Nov 8 2005;15(21):1961-1967.

14. Graaf MR, Riche! DJ, van Noorden CJ, Guchelaar HJ. Effects of statins and farnesyltransferase inhibitors on the

development and progression of cancer. Cancer Treat Rev. Nov 2004;30(7):609-641.

15. Kato K, Cox AD, Hisaka MM, Graham SM, Buss JE, Der CJ. Isoprenoid addition to Ras protein is the critical

modification for its membrane association and transforming activity. Proc Nat! Acad Sci U S A. Jul 15

1992;89(14):6403-6407.

16. de J ongh S, Ose L, Szamosi T, et al. Efficacy and safety of statin therapy in children with familial

hypercholesterolemia: a randomized, double-blind, placebo-controlled trial with simvastatin. Circulation. Oct 22

2002;1 06(17):2231-223 7.

17. National Institutes of Health Consensus Development Conference Statement: neurofibromatosis. Bethesda, Md.,

USA, July 13-15,1987. Neurofibromatosis. 1988;1(3):172-178.

18. Van Haasen PP, DeBruyn EEJ, Pijl YJ, et al. Wechslerlntelligence Scale for Chzfdren- Revised Nederlandstalige Uitgave. Lisse:

Swets & Zeitlinger B.V., The Netherlands; 1986.

C!HPTER 7 153

19. Van der Geest JN, Lagers-van Haselen GC, van Hagen JM, et al. Visual depth processing in Williams-Beuren

syndrome. Exp Brain Res. Oct 2005;166(2):200-209.

20. Baizer JS, Kralj-Hans I, Glickstein M. Cerebellar lesions and prism adaptation in macaque monkeys. J NettrophysioL

Apr 1999;81(4):1960-1965.

21. Pisella L, Rossetti Y, Michel C, et al. Ipsidirectional impairment of prism adaptation after unilateral lesion of anterior

cerebellum. Neurology. Jull2 2005;65(1):150-152.

22. Lezak lviD, Howieson DB, Loring DW, Hannay HJ, Fischer JS. Neurop{)lchological Assessment. 4th ed: New York:

Oxford University Press; 2004.

23. Medicines Evaluation Board. Summary ofproductcharacteristics ZCR-T-082004. The Hague, The Netherlands: Medicines

Evaluation Board; 2005.

24. Descheernaeker MJ, Ghesquiere P, Symons H, Fryns JP, Legius E. Behavioural, academic and neuropsychological

profile of normally gifted Neurofibromatosis type 1 children. J Intellect Disabil Res. Jan 2005;49(Pt 1):33-46.

25. Rappon LJ, Brines DB, Axelrod BN, Theisen ME. Full Scale IQ as Mediator of Practice Effects: The Rich Get

Richer. The Clinical Neurop{)lchologist. 1997;11 ( 4):3 7 5-380.

26. Struiksma AJM, Van Der Leij A, Vieijra JPM. Diagnostics tool of technical reading and basic spelling (Dutch version).

Amsterdam, The Netherlands: VU Publisher; 2004.

27. Ruijssenaars AJJM, Van Lieshout ECDM, Van Luit JEH. Mathematical problems and dyscalmlia (Dutch version) Rotterdam,

The Netherlands: Lemniscaat Publisher; 2004.

28. Brown RT, Amler RW, Freeman WS, et al. Treatment of attention-deficit/hyperactivity disorder: overview of the

evidence. Pediatrics. Jun 2005;115(6):e749-757.

154 THE NFl SIMVASTATTN TRLU.

, , '"fpe Blgetsma, PhD\ Lianne C. Ktab, MSo1.Z, Henriette A. Moll,,'MD PhD2

" ' ~ ~ ~ I ~

The NFJ CoRe Team (Cognitive Research Team)7 Erasmus NJC Univer.o;ity Medical Center­Sophia Children's Hot.piuU, Rotterdam, The Nethe1:lands, departments of

1Neuroscience and2 General Pediatrics.

JAMA, in press (2008)

Ct-L\PTER 7 I 1 5 5

LETTERSTOJAMA

Challenges for translational trials

To the editor: Krab et al.l investigated the possible benefit of statin therapy in children with

N eurofibmmatosis type 1 (NF1 ), a generic disorder which is known to be associated with learning

disabilicies.2 Using a randomized, double-blind, placebo-controlled trial it was found that simvastatin, a

3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitor, did not improve cognitive

function in children with NF1.1 The motivation for this clinical trial came from the beneficial effect of

Lovastatin (another HMG-CoA reductase inhibitor) on cognitive function in the NF1 +I- mouse model

NFL' These mice are heterozygous for a null mutation in neurofibromin (NF1 +I-), and exhibit

behavioral disorders that resemble those found in humans, and display deficits in physiological

correlates of memory.3

First of all, the mechanism for the effect of the statins, such as lovastatin and simvastatin,

on cognitive performance in the NF1 +I- mouse is not obvious.4 Statins are cholesterol-lowering drugs,

but neurofibromin is not involved in cholesterol metabolism, as it is a tumor suppressor protein.

Mutations in neurofibromin are associated with enhanced RAS activity, an oncogene that promotes

tumor growth and excessive intracellular signaling in neurons. It has been proposed that statins inhibit

the activation of RAS, effectively counterbalancing the enhanced RAS activity associated with

neurofibromin.3 The validity of this proposed mechanism remains to be verified.

Secondly, there are some remarkably differences in the preclinical trial by Li et al, and the

clinical trial by Krab et al Li et al used adult mice, whereas Krab et al selected children as participants

for their trial. Furthermore, a different statin therapy was applied. Lovastatin and simvastatin are HMG­

CoA reductase inhibitors with similar, though at some points different, pharmacokinecic proflles.5 It is

also unclear whether the authors considered the effects of nourishments, such as grapefruit juice, on

the pharmacokinetics of simvastatin.5 To what extend are age and pharmacokinetics critical?

The authors correctly conclude that based on the negative outcome of their trial,

simvastatin should not be prescribed to ameliorate coguicive deficits associated with NF1. It is difficult

to develop a productive and theoretically satisfactory animal model for NF1 because the factors that

engender the modeled symptoms or signs of the disorder are often imprecise and incomplete. I wonder

what the implications of the study by Krab et al. are for D the translational value of the current

experimental NF1 +I- mouse model, and ii) the benefit of statin therapy in general in NF1?

Jacobus F .A. Jansen, PhD

[email protected], Department of Medical Physics & Radiology, Memorial Sloan-Kettering Cancer

Center, New York, New York.

1. Krab LC, de Goede-Bolder A, Aarsen FK, et al Effect of simvaststin on Coguitive Functioning in Children With Neurofibromatosis Type 1: A Randomized Controlled Trial. JAMA. July 16, 2008 2008;300(3):287 -294.

2. Hyman SL, Shores A, Notth KN. The nature and frequency of coguitive deficits in children with neurofibromatosis type 1. Neurology. Oct 11 2005;65(7):1037-1044.

3. Li W, Cui Y, Kushner SA, et aL The HMG-CoA reductase inhibitor lovastatin reverses the learning and attention deficits in a mouse model of neurofibromatosis type 1. Curr Biol Nov 8 2005;15(21):1961-1967.

4. Johnston MV. Fresh ideas for treating developmental coguitive disorders.Curr Opin Ne11rol Apr 2006;19(2):115-118.

5. Neuvonen PJ, Backman JT, Niemi M Pharmacokinetic Comparison of the Potential Over-the-Counter Sratins simvastatin, Lovastatin, Fluvastatin and Pravastatin. Clin Pham1acokinet. 2008;47(7):463-474.

1 5 6 THE NF1 SlMVASTATIN TRTA.L

In Reply: Dr. Jansen raises questions about the rationale of our trial and the value of the Nf1 mouse

model for translational research. 1 It is important to re-emphasize that the NF1 protein is not involved

in cholesterol metabolism. NF1 is a negative regulator of RAS activity, and increased RAS signaling has

been shown to underlie the learning deficits of Njf+l- mice.2 The rationale to treat the cognitive deficits

with statins was based on 2 fundamental findings in the cancer literature: RAS requires post­

translational isoprenylation for proper signaling, and RAS transforming activity can be suppressed by

reducing the synthesis of the isoprenyl groups by statins.'.3,4 The molecular and behavioral deficits of

Njf+l- mice can indeed be ameliorated by decreasing RAS activity, either genetically or by

administrating farnesyl transferase inhibitors or statins.1•2

The pharmacokinetic profile of simvastatin was carefully considered in our study design.

Patients and their general practitioners were counseled to avoid the use of medication or food (such as

grapefruit juice) that could interfere with the cytochrome-P450-3A4 system in order to avoid significant

alterations in simvastatin blood levels.3 The treatment of central nervous system biochemical deficits

has unique considerations compared to treating hypercholesterolemia. For instance, hydrophilic HMG­

CoA inhibitors like pravastatin, fluvastatin, and atorvastatin show very limited penetration of the blood­

brain barrier.3 Our rationale to choose the highly lipophilic simvastatin rather than the very similar

lovastatin was based on the large amount of safety data available for the use of simvastatin in children. 5

Moreover, simvastatin but not lovastatin is approved for the treatment of familial hypercholesterolemia

for children in Europe. Our decision to conduct the trial using children with NFl, rather than adults,

was based largely on the well-characterized cognitive deficits of children with NFl. Additionally, early

intervention during childhood, the peak period of cognitive development, is likely to maximize the

benefits of treatment,

There are 3 major possibilities why the mouse findings could not be replicated in humans:

(1) simvastatin is not an adequate treatment of human NFl, (2) simvastatin is an efficacious treatment

but the current trial design used a treatment regimen (e.g., daily dose or length of trial) that was below

the clinically efficacious threshold, or (3) the Nf1 mouse model is inadequate. We consider the latter

possibility unlikely based on a large literature supporting the translational relevance of the NF1 mouse

model. The current data are particularly strong for modeling the cognitive aspects of NFl, as Nf1+1-

mice show learning and attention deficits in cognitive domains analogous to the patients. The human

brain is far more complex than a mouse brain, but it seems likely that the underlying molecular

mechanism that is being targeting is conserved across species. It is possible that statin treatment also

rescues the learning deficits in Njt flies.6 Further basic neuroscience and clinical research is therefore

needed to investigate how this knowledge can be translated to the patients.

Ype Elgersma, PhD, Lianne C Krab, MSc, Henriette A. Moll MD, PhD

[email protected], Erasmus MC University Medical Center, Rotterdam, the Netherlands.

1. Ii W, Cui Y, Kushner SA, et a!. The HMG-CoA reductase inhibitor lovasratin reverses the learning and attention deficits in a mouse model of neurofibromatosis type 1. Curr Bioi. 2005;15(21):1961-1967.

2. Costa RM, Federov NB, Kogan JH, et al. Mechanism for the learning deficits in a mouse model of neurofibromatosis type 1. Nature. 2002;415(6871):526-530.

3. Cban KK, Oza AM, Siu LL. The statins as anticancer agents. Clin Cancer Res. 2003;9(1):10-19. 4. Krab LC, Goorden SMI, Elgersma Y. Oncogenes on my mind: ERK and MTOR signaling in cognitive

diseases. Trends Genet. 2008;24(10):498-510. 5. de Jongh S, Ose L, Szamosi T, et al. Efficacy and safety of statin therapy in children witb familial

hypercholesterolemia: a randomized, double-blind, placebo-controlled trial with simvastatin. Circulation. 2002;106(17):2231-2237.

6, Vose L.R., Arora G.S., O'Brien PF, Hannan F. Pharmacologic Rescue of Behavioral Deficits in Drosophila NFl Mutants. NF Conference (abstract) 2008:p. 97,

CH,\l'TF.R 7 157

158 THE NFl SlMVASTATIN TRIAL

General discussion and future prospects

In this thesis, the impact of Neurofibromatosis type 1 (NFl) on daily life was investigated by

assessing neuropsychological functioning, school performance, quality of life, and motor

behavior in children with NFL Moreover, NF1-related neuroradiological abnormalities were

examined. In addition to characterizing specific NF1-related problems, we tried to identify

which aspects of NFl could be used as outcome measures when investigating potential

therapeutic interventions for NF1. In the last and major part of the thesis, several of the

outcome measures identified in our studies were incorporated in a randomized, double-blind,

placebo-controlled trial to investigate the effect of simvastatin on cognitive functioning in

children with NF1 (the NF1 simvastatin trial).

The studies described in this thesis are all carried out in a cohort of NF1 patients aged 7 to 16

years old who are attending the multidisciplinary outpatient clinic of the Erasmus MC - Sophia

Children's Hospital Rotterdam, the largest NF1 referral centre in the Netherlands. Most of the

data was gathered in the context of the NF1 simvastatin trial, which could potentially create a

bias in our data if the patients with the highest burden of cognitive problems would be more

inclined to participate than less affected patients. However, non-response analyses indicated this

was not the case. Therefore, the results of our studies can offer insight into the general NF1

patient population.

Baseline assessment

School performance

In chapter 3, we revealed that NF1 has a large impact on school performance. An important

question outstanding is whether we can predict problems in school performance based on

scores in specific domains of the neuropsychological profile of NF1 patients. Ideally, we would

want to identify risk factors for school problems at a young age. This would enable us to

provide parents and children with a reliable prognosis of school performance, but would also

facilitate early intervention in the hopes of preventing school problems later on. Predictive

factors can for instance be identified by using regression analysis on data obtained from

longitudinal studies, in which children undergo detailed neuropsychological testing at preschool

age and quantitative assessment of school performance at an older age. However, such a

longitudinal study requires a long follow-up period. As a short-term alternative, we can use the

160 GENERAL DISCUSSION .'\.ND FUTURE PROSPECTS

cross-sectional data from chapter 3 to explore the relationship between neuropsychological

functioning and school performance.

The results of this preliminary analysis are displayed in supplementary table 1 at the end of the

discussion. The high correlations between learning efficacies and IQ can be explained by the

fact that more than half of the learning disabilities in NF1 occur in the context of a lowered IQ.

In addition, we observed a link between performance on some, but not all, of the

neuropsychological tests frequently affected in NF1 and school scores, which indicates these

neuropsychological tests might be relevant outcome measures that can be used in studies

investigating the effect of potential treatments for cognitive problems in NF1. The lack of a

clear association between attention deficits and school performance is unexpected, as a

relationship between ADHD and literacy based learning disabilities is suggested in NF1 patients

as well as in the general population.!· 2 Due to the limited sample size available for these

analyses, the correlation of a combination of individual factors with school scores can only be

estimated from data of larger studies.

After identifying predictive factors for school performance in longitudinal studies, an important

follow up question is whether school problems can actually be reduced, and if so, to what

extent, by targeting deficits in these predictive domains with either (tailored) remedial teaching

programs or drug therapies. The ideal method to study these effects is by performing a

randomized trial. However, as remedial teaching is often already implemented in regular care, it

is not expected that parents will participate in a randomized trial which incorporates a control

group without remedial teaching. Therefore, one would have to refer to before- and after

measurements.

Health Related Quality of Life

In chapter 4, we observed that NF1 children, in contrast to their parents, report difficulties in

only one domain of Quality of Life of the Child Health Questionnaire (CHQ). Although a

discrepancy between ratings of parents and children does not imply one of them is more

accurate,3 these results are intriguing. Only one other group has investigated child self-reports in

NF1 so far, and this study showed children do report significant problems on multiple domains

of the TNO-AZL Quality of Life questionnaire (TACQOL).4 The discrepancy between our

study and that of Graf et al. could be caused by (cultural) differences between patient

populations, but also by methodological distinctions between the questionnaires used. The

CHQ assesses the prevalence of problems, whereas scores on the TACQOL are determined by

CHAPTERS 161

the amount of distress related to these problems regardless of the prevalence. Possibly, children

with NF1 cannot accurately perceive or report either one of these aspects, which might be

related to the 'positive illusory bias' mentioned in the introduction.s-7 A way to investigate

whether methodological differences can explain the differences in problem scores across

questionnaires, and to investigate whether this sensitivity to the way questions are posed is NF1-

specific, would be to administer both the CHQ and the TACQOL questionnaires to both our

study population and to a group of healthy control children and compare scores.

We revealed that parent-reported Health Related Quality of Life scores are strongly sensitive to

behavioral problems (rated by teachers). This insight suggests an opportunity to influence not

only behavior itself but also quality of life in general by addressing these behavioral problems,

for instance with specific medication, behavioral training programs or family education. A

problem with this approach is that the children with NF1 themselves did not seem to notice

behavioral problems, as they scored their own behavior as significantly better than average. This

discrepancy could again be explained by the positive illusory bias. It is not clear whether

standard behavioral training programs would be beneficial for the self-perception of children

with NF1. One can imagine that confronting children with a positive illusory bias with their

inadequate behavior could even have a negative rather than positive impact on their self-esteem

or quality of life, which calls for a careful approach.

From our study on school performance we can conclude that school problems pose a major

burden for parents and children, as children frequently need multiple types of remedial teaching

and often have to repeat a grade. However, an unexpected finding of our study on Quality of

Life was that these school problems do not seem to contribute to Health Related Quality of Life

scores. It could be that our school scale was not sensitive enough to pick up a relationship

between school problems and Health Related Quality of Life scores. Another possible

explanation is that the parent's percept of their child's school performance is not linearly related

to their child's objective academic achievement. Parents of a child that is thriving in a special

education class could experience less school-related concerns than parents of a child struggling

to keep up with regular education. It would be interesting to incorporate parental opinions of

their child's school performance as a covariate in future studies.

Motor functioning

In chapter 5, we showed that children with NF1 display deficits on fine motor functioning and

visual integration, and adaptation of hand movements to prism glass distortion. Our study did

162 GENERAL DISCUSSION AND FUTURE PROSPECTS

not reveal a specific anatomical correlate of these motor problems in NFL By making use of

mouse models for NF1, future studies could assess the functional integrity of brain areas

involved in motor behavior. In order to determine whether cerebellar function is affected in

NF1 we could assess cerebellar synaptic plasticity, for instance in the GABA-agric Purkinje cell

/ deep cerebellar nuclei synapses. Also, Nf1 mice could be tested with cerebellum-specific

motor adaptation tasks, such as conditioning of eyeblink responses.8 The benefit of eyeblink

conditioning is that it can be assessed in mice as well as patients.9. 10 If Nf1 mice and humans

turn out to have parallel deficits in eyeblink conditioning, this test can serve as a unique

translational outcome parameter in future studies assessing the effect of potential drug therapies

on motor functioning in Nf1 mice and NF1 patients.

Unidentified Bright Objects

Our study on Magnetic Resonance (MR) abnormalities in NF1 has gained insight into the nature

of T2-weighed hyperintensities (chapter 6). However, the effect of these Unidentified Brain

Objects (UBOs) on brain functioning and cognition is still unclear. Previous studies

investigating the relationship between UBOs and cognition mostly identified UBOs visually on

conventional T2-weighed MR images,ll-19 However, as we and other studies showed that

Apparent Diffusion Coefficients (ADC values) are also elevated in normal-appearing brain areas

in NF1 patients,20-23 visual identification might not be an accurate parameter for brain pathology

in NFL The information gathered in our studies allows for a more detailed investigation of the

relationship between NF1-specific brain pathology and cognition.

Preliminary analysis of our data is shown in supplementary table 1. Although there is little

correlation between average ADC-value and neuropsychological performance, there does seem

to be a tendency for a positive relationship between average ADC value and learning efficacies for

technical reading, comprehensive reading and spelling. Although these results should be viewed

in the light of the small study group and the large amount of statistical comparisons made, it

must be noted that only one out of the 16 comparisons made for the neuropsychological tests

has a negative correlation coefficient. If anything, our data suggest a positive relationship

between ADC values and cognitive functioning, with higher ADC values (which are indicative

of UBOs) related to higher test scores. This is opposite to all previous reports in literature, and

there is no straightforward explanation for this. One would imagine that myelin disturbances

impair axonal conductance, and lead to inadequate neuronal signaling and impaired cognitive

functioning. It is tempting to speculate that UBOs might be indicators of a compensatory

mechanism to reduce increased neuronal inhibition. NF1 heterozygous knockout mice do not

OL\PTERB 163

seem to display UBOs,24 which makes it difficult to assess the effect of UBOs on neuronal

functioning.

Potential outcome measures: prism adaptation and ADC values

Part of the reason we conducted the neurophysiological and neuroradiolocical studies described

in chapter 5 and 6 was to identify possible objective, placebo-insensitive outcome measures to

assess the effect of potential therapies on cognitive functioning in NF1 patients.

Performance on the prism adaptation test could potentially be more sensitive to changes in

neuronal functioning than neuropsychological tests for higher order cognitive skills, because the

task depends upon a limited neuronal circuit,25 and deficits can not easily be compensated for.26,

27 Although some children were excluded because they did not understand or adhere to task

instructions or were left-handed, overall the prism adaptation test is easy to perform,28 rapid,

and can be quantified objectively. As adaptation of hand movements to prism glass distortion

was found to be impaired in NF1 patients, we decided to incorporate this test as an objective

outcome measure in our clinical trial.

As described in chapter 6, ADC values were significandy higher in children with NF1 than in

controls. In addition, we showed that the measurement of ADC values was reproducible, and

NF1 children were able to undergo MR investigation without needing sedation. Previous studies

indicate that UBOs can resolve over time, and that ADC values are more sensitive to brain

pathology in NF1 than UBOs.21. 29. 30 Thus, it was conceivable that if statins reduce brain

pathology in NF1, this could be picked up by measuring ADC values. Therefore, we

incorporated ADC values as an objective outcome measure in our trial.

Investigating the effect of statins on cognition in children with NF1

In chapter 2, we reviewed the insights in the molecular and cellular mechanisms underlying the

cognitive deficits in NF1 and affiliated diseases among the neuro-cardio-facial-cutaneous and

Hamartoma syndromes. Research in mouse mutants has revealed that the cognitive deficits of

these diseases evolve around elevated activity of the RAS/ERK and RAS/PI3K/MTOR

pathways, which leads to changes in synaptic plasticity. RAS activity can be decreased by

attacking it's Achilles' heel: its requirement to be isoprenylated.31 Statins can decrease the

164 GENERU DISCUSSION AND FUTURE PROSPECTS

production of isoprenyl groups by inhibiting HMG-CoA reductase, the rate-limiting enzyme in

the mevalonate synthesis pathway. Statins are cholesterol-lowering drugs, used by millions of

people worldwide, and show very good safety profiles in adults and children.32, 33 A

breakthrough in the pursuit of a treatment for the cognitive deficits in NF1 patients was made

when it was discovered that short-term lovastatin treatment could reduce RAS activity, and

rescue the deficits in synaptic plasticity, learning, memory and attention in Nf1 rnice.34 The

favorable safety profile of statins offers a unique opportunity to translate these preclinical

findings and to assess the effect of a targeted treatment on cognitive function in NF1 patients.

In chapter 7, we report the results of the first randomized, placebo-controlled, double blind trial

to assess he effect of statins on cognitive function in children with NFL Sixty-two children with

NF1 aged 8 to 16 years were treated with simvastatin or placebo once a day for 12 weeks. The

effect of simvastatin was assessed using neuropsychological, neurophysiological and

neuroradiological outcome parameters, carefully selected from the tests found to be impaired in

our NF1 patient population in the baseline studies performed in chapters 3, 5 and 6.

We did not observe an effect of simvastatin on the primary outcome measures (Rey Complex

Figure test, Cancellation test [speed], Prism Adaptation and average brain ADC-value). On the

secondary outcome measures, we found a significant improvement in the simvastatin group in

object assembly scores (B=0.54, CI: 0.08-1.01), which was specifically observed in children with

poor baseline performance (B=O.SO, CI: 0.29-1.30).

The changes in object assembly found in our study could be a spurious finding, and need to be

verified in other studies. An important question is what are the consequences of improvement

in object assembly for daily life. Object assembly is postulated to assess visual analysis (the

ability to synthesize an image from fragmented visual information),3S which is an important

prerequisite ability for reading and spelling, and is used in advanced mathematical problem

solving.36· 37 Unfortunately, preliminary analysis of our data on school performance does not

reveal any correlation between performance on object assembly and learning efficacies on

technical reading, comprehensive reading, spelling or mathematics in children with NF1 (all R

0.0-0.1). Thus, we cannot say whether improvements in object assembly would be beneficial to

children with NF1 on the long run. Another concern is that our trial included outcome

measures that tap into some of the other skills required for performance on the object assembly

task, such as visual synthesis (Block Design test) and visual motor coordination (Beery VMI

test), but these were not changed in our trial.

CJ-L\PTER 8 165

The overall conclusion of our trial was that 12-week simvastatin treatment does not improve

cognitive functioning in NF1 patients.

Factors confounding trial outcome

Several factors could have attributed to the negative outcome of the NF1 simvastatin trial.

First, the treatment duration used in our study might have been too short. We based the length

of our trial on the observation that statin treatment normalized the plasticity impairment and

cognitive phenotype of Nf1 mice within days,34 and the reports that clinically significant

reduction of cognitive problems in children can be reached within days to weeks (for instance in

the treatment of attention deficits in ADHD, reviewed by Brown et al 38). However, since

precedents for translational trials into cognition are rare, we cannot exclude the possibility that

the effect of simvastatin on higher cognitive functions in humans would require a longer

treatment period than 12 weeks.

Second, relatively little is known about the comparison of pharmacokinetic and

pharmacodynamic findings in mice and humans. It is conceivable that the therapeutic effect of

simvastatin on human brain function was hampered by suboptimal availability of the drug or

due to inefficient crossing of the blood brain barrier.

With respect to the availability of the drug, it is known that statins undergo a large first pass

effect.39 Statins are generally administered orally, as an inactive lactone prodrug that is converted

into an active hydroxyacid form by carboxylesterases. This conversion is much more rapid in

rodents than in humans,40, 41 which potentially results in differences in the availability of the

drug. In humans, the inhibition of HMG-CoA reductase in plasma reaches a peak after a few

hours, and within approximately 8 hours fall back to baseline levels.39 It is unclear if this

relatively short time interval is sufficient to reach a sustained effect on protein isoprenylation. In

our trial, we administered simvastatin in the morning in order to time the peak of statin activity

during school hours.

The way statins penetrate the blood brain barrier is still unclear, and does not solely depend

upon the lipophilicity of the drug. The lactone and acid forms are shown to interact differently

with efflux and uptake transporters present in the blood brain barrier.42 The brain penetration of

the lactone form, but not the active acid form, is limited by the drug transporter P-glycoprotein.

166 GENER.U DISCUSSION AND FUTURE PROSPECTS

Despite this, only a limited amount of the active acid form seems to penetrate the blood brain

barrier.43 Brain carboxylesterases can convert the lactone form to the acid form, but conversion

does not seem to take place the other way around. It is unclear which form of statins is the most

important inhibitor of brain HMG-CoA reductase. Part of the studies in the paper of U et al.

were performed using subcutaneous injections of the acid form (the RAS activity assays, the

plasticity experiments and the task for learning and memory).34 Differences in the way statins are

administered might influence the effect on the mevalonate pathway in the brain.

Regardless of the way statins reach the brain, statin treatment has been shown to reduce brain

cholesterol synthesis in mice,44 and thus statins potentially affects isoprenyl concentrations in

the neurons. However, their influence on brain cholesterol turnover in hyperlipidemic patients is

unclear.45, 46 The effect of simvastatin on brain cholesterol synthesis was not assessed in our

study but could be determined by measuring 24(S)-hydroxycholesterol, a serum marker for brain

cholesterol metabolism that is suggested to be responsive to simvastatin treatment.45, 47

Third, we observed large improvements on the neuropsychological outcome measures in both

treatment groups, which might be attributable to a placebo-effect,48 but also to practice effects

resulting from repeated assessment with the same tests. Practice effects can be observed even

when applying parallel versions of a test.49 Although the placebo-controlled design of our trial

should account for these effects, the mean score on 3 out of 9 neuropsychological outcome

measures was increased up to a normal score in the placebo group after 12 weeks. Children in

the simvastatin group would have had to improve beyond the normative average before a

treatment effect could have been identified. As statins do not seem to improve cognitive

function in NFl and wild type mice beyond normal,34 it is possible that a performance ceiling

was reached for these measures, which might have hampered identification of additional effects

of simvastatin.

Finally, it cannot be excluded that the outcome measures used in our study were too specific to

pick up overall improvements in daily life functioning. Potentially, subde changes in multiple

cognitive domains could together result in substantial changes in daily life functioning and

subjective well-being without significandy improved scores on tests for the separate cognitive

skills.

C!HPTER K 167

Recommendations for future trials

Despite the fact that our NF1 simvastatin trial did not provide evidence that short-term

simvastatin treatment has an effect on cognitive functioning in NF1 patients, the favorable

safety profile of statins, and the limitations of this first trial, do call for further randomized

double-blind placebo-controlled trials. Below, we discuss some important recommendations for

the design of these future studies.

Treatment period

The major recommendation for follow-up studies is to treat children for a much longer period.

A longer treatment duration in future trials may attenuate the placebo effect, wear off practice

effects, enable a longer exposure to the highest therapeutic dose, and allow for a longer time for

the brain to restore function. Moreover, a longer treatment duration would allow inclusion of

real-life outcome measures such as school performance, behavior and quality of life. A follow­

up trial with a treatment period of 1 year is currently in preparation at the Erasmus MC - Sophia

Children's Hospital Rotterdam.

Treatment dose

Increasing the therapeutic dose does not seem desirable due to the lack of safety studies in

children with higher doses, and the increasing risk of side effects observed in adults.50 In

addition, the effect of simvastatin on low-density lipoprotein cholesterol at 12 weeks was similar

to the decrease achieved after 48 weeks of simvastatin treatment in a previous pediatric study.32

This indicates that, at least in the liver, the treatment dose was optimal with respect to inhibition

of the mevalonate pathway. Moreover, we did not observe any relationship between the dose of

simvastatin and decrease of low-density lipoprotein cholesterol (chapter 7), or the change in

object assembly score (unpublished observations).

Patient selection

It might be recommended to select individual patients based on baseline impairments, because

our subgroup analysis suggests that patients with low baseline scores might respond better to

simvastatin therapy. Excluding children with normal scores would also decrease the change of

children reaching a ceiling in their performance. However, drawbacks of patient selection are

that it introduces a selection bias, which lowers the generalisability of the study, and limits the

number of eligible patients, especially when selecting for performance deficits in multiple

outcome measures.

168 GENERAL DISCUSSION AND FllTURE PROSPECTS

Although the diagnosis of NFl is still based only on clinical criteria, it does seem important to

include patients only if the NFl diagnosis has been confirmed with genetic testing. Despite the

fact that patients with affiliated disorders in the RAS pathway might also respond to statin

treatment, the magnitude of response might differ according to the specific gene affected, which

would add to the variability in the study. The fact that even with thorough clinical assessment,

patients with other disorders can erroneously receive an NFl diagnosis is illustrated by the

unexpected finding of a PTPNll T468M point mutation, indicative of Noonan Syndrome or

LEOPARD syndrome, 51,52 in one of the participants of the NFl simvastatin trial.

Selection of outcome measures

In general, it is recommended to select a limited amount of primary outcome measures, in order

to prevent having to correct statistical findings. Outcome measures used in future studies need

to have a good test-retest reliability, as this is a major determinant of the power of the study. It

is of pivotal importance to select tests on which NF1 patients show impairments (although the

exact cut-off for the level of impairment can be adjusted), in order to have room for

improvement and to prevent ceiling effects.

Another important recommendation is to select outcome measures that have a high predictive

validity, so that changes on these tests can be interpreted in terms of their consequences for

daily life functioning. Our preliminary analysis indicates that some neuropsychological tests are

more suitable in this respect than others. For instance, it might be recommended to include total

IQ and neuropsychological tests for language, but to drop the Judgment of Line Orientation

Test as it does not seem to correlate to performance in any academic domain. Although object

assembly does not seem to predict school performance in our population, we do recommend

incorporating it as an outcome measure in future trials, in order to examine whether our current

results can be replicated, and to be able to compare results across studies.

As indicated above, a longer treatment duration would enable the assessment of the effect of

statins on daily life functioning. Considering the problems identified in our NFl patient

population in our studies on school performance and quality of life, we recommend to include

detailed quantitative assessment of academic achievement, as well as standardized validated

questionnaires on behavior, quality of life, and also self esteem, from the perspective of parents,

teachers and children.

CRWrER8 169

Behavior was assessed in our trial, but was not selected as an outcome measure because scores

at baseline were less than 1 SD below normative values. Our results indicate that self-reported

behavior might be sensitive to simvastatin treatment, as children's scores on a validated

behavioral questionnaire (the Youth Self Report Form, Achenbach) were marginally significandy

more improved in the simvastatin group (p=0.06 on multivariate analysis). Parent and teacher's

ratings revealed no significant differences between the simvastatin and placebo groups, although

in both groups the scores improved substantially. Because the number of children that were old

enough to fill out this questionnaire (>11 years) and returned their forms at baseline as well as

after 12 weeks was small, these results should be interpreted with care, but do stress the

importance of including behavioral questionnaires in future studies.

Based on progressive insight gathered in our trial, we do not recommend incorporating the

prism adaptation task in future trials. On repeated assessment, the missing data on this test were

relatively large. Also, when closely examining the prism adaptation data in the placebo group, it

seems that performance on this task is not as stable as we had anticipated. On the group level,

the distribution of children over the 'adapting' and 'not-adapting' categories was similar at

baseline and after 12 weeks. However, on an individual level, performance is very variable, with

only 13 out of 26 children falling into the same category at both time points.

Although 12 weeks simvastatin therapy did not have a significant effect on mean brain ADC

values, this could be confounded by the short treatment duration. However, the potential

benefits of incorporating the objective investigation of brain pathology in future trials should be

balanced against high costs and time-consuming nature of this measurement, and its low

predictive validity for school problems and neuropsychological problems.

Future trials are strongly recommended to incorporate determination of 24(S)­

hydroxycholesterol levels in serum to assess the effect of simvastatin on brain cholesterol

synthesis.

Other applications of statins

Applications of statins for other NF1-related problems

Recent studies reveal other potential therapeutic options for statins in NF1 besides treatment of

cognitive deficits. Oral lovastatin treatment can rescue the delayed bone repair of mice with

170 GENERAL DISCUSSION AND FUTURE PROSPECTS

conditional bi-allelic inactivation of neurofibromin in the developing limbs and cranium,

probably via repression of ERI<:.-activity.53 In addition, in NFl Malignant Peripheral Nerve

Sheet Tumor cell lines, lovastatin and a farnesyl transferase inhibitor were found to have a

synergistic effect, and together reduced RAS isoprenylation, decreased cell proliferation and

induced apoptosis. Single administration of either agent did not have this effect.54 It would be

very interesting to assess the effect of statins on bone repair and Malignant Peripheral Nerve

Sheet Tumors in clinical trials. However, as these complications of NFl is rare, they need to be

investigated in separate studies.

Applications of statins for cognitive impairments in other diseases

As reviewed in chapter 2, statins do not only offer prospects of developing a treatment for NF1.

Statins are also of great interest for other diseases in the RAS pathway such as the Neuro­

Cardio-Facial-Cutaneous syndromes. In addition, statins could potentially target the molecular

disturbances underlying the Hamartoma syndromes, which are not only co-regulated by RAS,

but also critically dependent upon RHEB, another member of the RAS family that requires

isoprenylation. However, awaiting further clinical trials, a first step should be to obtain a proof

of principle in the preclinical models of these diseases.

The potential effect of statins on for instance Alzheimer's disease, Multiple Sclerosis, and the

incidence of stroke is of great interest to the general population.33 It is also tempting to

speculate whether statins could improve cognitive functioning in otherwise healthy humans.

This seems unlikely, as statins do not alter the learning phenotype of wild type mice or flies.34. ss

However, in another study, in acute application of statins to brain slices of wild-type mice did

not seem to affect long term potentiation. 56 The reports on the effect of statins on cognition in

humans are conflicting as well. Although several case reports indicate statins have a negative

effect on cognition,s7 results from clinical trials in patients with hypercholesterolemia and mild

cognitive decline are controversial. Some of these trial indicate that statins lead to relatively

lower learning capacities, sa, 59 but others do not find an effect, 60, 61 or report a favorable effect

on cognitive decline.62 In all, it seems that the effect of statins on the brain will remain a hot

topic for quite some time.

CHAPTERR 171

Conclusion

Our studies indicate that NF1 has a large impact on daily life functioning, as NF1 children show

substantial problems in school performance and motor functioning and, at least according to

parents, have a lower quality of life. Awareness of the problems associated with NF1 may

facilitate timely recognition and appropriate intervention. Our studies have pointed out several

potential targets for structural support, such as behavioral problems. Also, we have identified

several outcome measures that can be used to assess potential treatments for cognitive deficits

in NF1. Although the preclinical studies were promising, short-term statin treatment did not

improve cognitive functioning. Still, further clinical trials are needed to reveal whether long-term

statin treatment can improve school performance, behavioral problems, and quality of life in

children with NF1.

Supplementary table 1. Correlations between Learning efficacies and performance on Neuropsychological tests

and average ADC-value of the brain.o.b

Didactic domain' Technical Reading

Comprehensive Reading Spelling Mathematics

Neuropsychological domain IQ Memory

Verbal shon termb V erballong term Non-verbal

Language E>qJresslve Receptive

Visual spatial skills Visual integration Visual motor integration JLO Executive skills Rote memory Divided attention Verbal fluency Concept formation Preservations

Attention Sustained Selective

Neuroradiological parameter

0.5

0.3 0.4

0.4 0.4

0.3 0.3

0.8

0.5 0.6

0.5 0.5

0.4 0.4

0.3

0.5

0.3 0.5

0.5 0.4

0.4

Average ADC-value 0.3 0.3 0.4

0.7

0.3 0.5

0.4 0.6

0.4 0.4 0.3

'Values represent Pearson's R. Only significant correlations (p<O.OS) are displayed;'-' indicates no significant correlation.

bThe average ADC-value did not correlate significandy with any neuropsychological test except for verbal shan term

memory (R=0.3).

'N = 39 to 51, depending upon the availability of the didactic information and neuroradiological assessment.

JLO: Judgment of Line Orientation Test; ADC: Apparent Diffusion Coefficient.

172 GENERAL DISCUSSION AND FUTURE PROSPECTS

References

1. Hyman SL, Arthur Shores E, North KN. Learning disabilities in children with neurofibromatosis type 1: subtypes,

cognitive profile, and attention-deficit-hyperactivity disorder. Dev Med Child Neurol2006;48:973-7.

2. Keulers EH, Hendriksen JG, Peron FJ, Wassenberg R, Wuisman-Frerker MG, Jolles J, et al. Methylphenidate

improves reading performance in children with attention deficit hyperactivity disorder and comorbid dyslexia: An

unblinded clinical trial. Eur J Paediatr Neurol2007;11:21-8.

3. Eiser C, Morse R. Can parents rate their child's health-related quality oflife? Results of a systematic review. Qual Life

Res 2001;10:347-57.

4. Graf A, Landolt MA, Mori AC, Boltshauser E. Quality of life and psychological adjustment in children and

adolescents with neurofibromatosis type 1. J Pediatr 2006;149:348-53.

5. Barton B, North K. The self-concept of children and adolescents with neurofibromatosis type 1. Child Care Health

Dev 2007;33:401-8.

6. Owens JS, Hoza B. The role of inattention and hyperactivity/impulsivity in the positive illusory bias. J Consult Clin

Psychol2003;71:680-91.

7. Heath NL, Glen T. Positive illusory bias and the self-protective hypothesis in children with learning disabilities. J

Clin Child Adolesc Psychol2005;34:272-81.

8. Koekkoek SK, Hulscher HC, Dortland BR, Hensbroek RA, Elgersma Y, Ruigrok TJ, et al. Cerebellar LTD and

learning-dependent timing of conditioned eyelid responses. Science 2003;301:1736-9.

9. Koekkoek SK, Yamaguchi K, Nfil.ojkovic BA, Dortland BR, Ruigrok TJ, Maex R, et al. Deletion ofFMR1 in Purkinje

cells enhances parallel fiber LTD, enlarges spines, and attenuates cerebellar eyelid conditioning in Fragile X

syndrome. Neuron 2005;47:339-52.

10. Smit AE, van der Geest JN, Vellema M, Koekkoek B, Willemsen R, Govaerts LC, et al. Savings and Extinction of

Conditioned Eye blink Responses in Fragile X Syndrome. Genes Brain Behav 2008.

11. Feldmann R, Denecke J, Grenzebach M, Schuierer G, Weglage J. Neurofibromatosis type 1: motor and cognitive

function and T2-weighted MRI hyperintensities. Neurology 2003;61:1725-8.

12. North K, Joy P, Yuille D, Cocks N, Mobbs E, Hutchins P, et al. Specific learning disability in children with

neurofibromatosis type 1: significance ofMRI abnormalities. Neurology 1994;44:878-83.

13. Goh WH, I<hong PL, Leung CS, Wong VC. T2-weighted hyperintensities (unidentified bright objects) in children

with neurofibromatosis 1: their impact on cognitive function. J Child Neurol2004;19:853-8.

14. Moore BD, Slopis JM, Schomer D, Jackson EF, Levy BM. Neuropsychological significance of areas of high signal

intensity on brain MRis of children with neurofibromatosis. Neurology 1996;46:1660-8.

15. Denckla MB, Hofman K, Mazzocco MM, Melhem E, Reiss AL, Bryan RN, et al. Relationship between T2-weighted

hyperintensities (unidentified bright objects) and lower IQs in children with neurofibromatosis-1. AmJ Med Genet

1996;67:98-102.

16. Hyman SL, Gill DS, Shores EA, Steinberg A, North KN. T2 hyperintensities in children with neurofibromatosis type

1 and their relationship to cognitive functioning. J Neurol Neurosurg Psychiatry 2007;78:1 088-91.

17. Hyman SL, Gill DS, Shores EA, Steinberg A, Joy P, Gibikote SV, et al. Natural history of cognitive deficits and their

relationship to MRI T2-hyperintensities in NFL Neurology 2003;60:1139-45.

18. Perner RE, Chaudhuri R, Bingham], Cox T, Hughes RA. MRI in neurofibromatosis 1. The nature and evolution of

increased intensity T2 weighted lesions and their relationship to intellectual impairment. J Neurol Neurosurg

Psychiatry 1993;56:492-5.

CH,\l'TER8 173

19. Legius E, Descheemaeker MJ, Steyaen J, Spaepen A, Vlietinck R, Casaer P, et al. Neurofibromatosis type 1 in

childhood: correlation ofMRl findings with intelligence. J Neural Neurosurg Psychiatry 1995;59:638-40.

20. Alkan A, Sigirci A, Kudu R, Ozcan H, Erdem G, Aslan M, et al. Neurofibromatosis type 1: diffusion weighted

imaging findings of brain. Eur J Radiol2005;56:229-34.

21. Tognini G, Ferrozzi F, Garlaschi G, Piazza P, Patti A, Virdis R, et al. Brain apparent diffusion coefficient evaluation

in pediatric patients with neurofibromatosis type 1. J Comput Assist Tomogr 2005;29:298-304.

22. Sheikh SF, Kubal WS, Anderson AW, Mutalik P. Longitudinal evaluation of apparent diffusion coefficient in

children with neurofibromatosis type 1. J Comput Assist Tomogr 2003;27:681-6.

23. Eastwood JD, Fiorella DJ, MacFall JF, Delong DM, Provenzale JM, Greenwood RS. Increased brain apparent

diffusion coefficient in children with neurofibromatosis type 1. Radiology 2001;219:354-8.

24. Rosenbaum T, Engelbrecht V, I<rolls W, van Dorsten FA, Hoehn-Berlage M, Lenard HG. lviRI abnormalities in

neurofibromatosis type 1 (NF1): a study of men and mice. Brain Dev 1999;21:268-73.

25. Hopp JJ, Fuchs AF. The characteristics and neuronal substrate of saccadic eye movement plasticity. Prog Neurobiol

2004;72:27 -53.

26. Pisella L, Rossetti Y, l'V!ichel C, Rode G, Boisson D, Pelisson D, et al. Ipsidirectional impairment of prism adaptation

after unilateral lesion of anterior cerebellum. Neurology 2005;65:150-2.

27. Straube A, Deubel H, DitterichJ, EggenT. Cerebellar lesions impair rapid saccade amplitude adaptation. Neurology

2001;57:2105-8.

28. Vander GeestJN, Lagers-van Haselen GC, van HagenJM, Brenner E, Govaerts LC, de Coo IF, et al. Visual depth

processing in Williams-Beuren syndrome. Exp Brain Res 2005;166:200-9.

29. I<raut MA, Gerring JP, Cooper KL, Thompson RE, Denckla MB, Kaufmann WE. Longitudinal evolution of

unidentified bright objects in children with neurofibromatosis-1. Am J Med Genet A 2004;129:113-9.

30. Imamura A, Matsuo N, Ol..-uda M, Morita H, Iwata M, Yamazaki Y, et al. Serial MR imaging and (1)H-MR

spectroscopy of unidentified bright objects in a case of neurofibromatosis type 1. Brain Dev 2005.

31. Konstantinopoulos PA, Kararnouzis MV, Papavassiliou AG. Post-translational modifications and regulation of the

RAS superfamily of GTPases as anticancer targets. Nat Rev Drug Discov 2007;6:541-55.

32. de J ongh S, Ose L, Szamosi T, Gagne C, Lambert M, Scott R, et al. Efficacy and safety of statio therapy in children

with familial hypercholesterolemia: a randomized, double-blind, placebo-controlled trial with simvastatin. Circulation

2002;106:2231-7.

33. Menge T, Hartung HP, Stuve 0. Statins--a cure-all for the brain? Nat Rev Neurosci 2005;6:325-31.

34. Li W, Cui Y, Kushner SA, Brown RA, Jentsch JD, Frankland PW, et al. The HMG-CoA reductase inhibitor

lovastatin reverses the learning and attention deficits in a mouse model of neurofibromatosis type 1. Curr Bioi

2005;15:1961-7.

35. De Bruyn EEJ VdSG, Van Haasen, PP. Wechsler Intelligence Scale for Children - Revised, Dutch edition

(verantwoording). Lisse: Swets & Zeidinger; 1986.

36. Struiksma AJM, Van Der Leij A, Vieijra JPM. Diagnostics Tool of Technical Reading and Basic Spelling (Dutch

version) Amsterdam, the Netherlands: VU Publisher; 2004.

37. Ruijssenaars AJJM, Van Lieshout ECDM, Van LuitJEH. Mathematical Problems and Dyscalculia (Dutch version)

Rotterdam, the Netherlands: Lemniscaat Publisher; 2004.

38. Brown RT, Amler RW, Freeman WS, Perrin JM, Stein MT, Feldman HM, et al. Treatment of attention­

deficit/hyperactivity disorder: overview of the evidence. Pediatrics 2005;115:e7 49-57.

39. Pentikainen PJ, Saraheimo M, SchwartZ JI, Amin RD, SchwartZ MS, Brunner-Ferber F, et al. Comparative

pharmacokinetics of lovastatin, simvastatin and pravastatin in humans. J Clin Pharmacal 1992;32:136-40.

174 GENERAL DISCUSSION AND FUTURE PROSPECTS

196 PERSONAL NOTES

PUBLISHED ABSTRACTS

1. Lianne C. Krab, Arja De Goede-Bolder, Femke K. Aarsen, Saskia M.P. Pluijm, Marlies J. Bouman, Jos N. van der Geest, Maarten Lequin, Coriene E. Catsman-Berrevoets, Willem

Frans M. Arts, Steven A. Kushner, Alcino J. Silva, Chris I. De Zeeuw, Henriette A. Moll,

Ype Elgersma. The effect of simvastatin on cognitive functioning in children with

Neurofibromatosis type 1: a randomized, double-blind, placebo-controlled trial. 1Jth

European Neurofibromatosis Meeting, Kil!arnry, Ireland, 2008 (oral presentation).

2. Lianne C. Krab, R. Oostenbrink, Arja de Goede-Bolder, Femke K. Aarsen, Ype Elgersma,

Henriette A. Moll. Health Related Quality of Life in children with Neurofibromatosis Type

1 - Contribution of demographic factors, disease related factors, and behavior. 131h European

Neurofibromatosis Meeting, Killarnry, Ireland, 2008 (oral presentation).

3. L.C. Krab, F.K. Aarsen, A. de Goede-Bolder, C.E. Catsman Berrevoets, W.F. Arts, H.A.

Moll, Y. Elgersma. Impact of Neurofibromatosis type 1 on school performance. 2007

Neurofibromatosis Conference: Models, mechanisms and therapeutic targets, Park Ciry, Utah, USA,

2001 (poster presentation)

4. L.C. Krab, F.K. Aarsen, A. de Goede-Bolder, C. Catsman-Berrevoets, H.A. Moll, Y.

Elgersma. Impact of Neurofibromatosis type 1 (NF1) on school performance. 12th

European Neurofibromatosis Meeting, Lisbon, Portuga~ 2001 (oral presentation).

5. Lianne C. Krab, MScl,z, Rianne Oostenbrink, MD PhD2, Femke K. Aarsen, MA3, Arja de

Goede-Bolder, Coriene E. Catsman MD PhD3, Henriette A. Moll, MD PhD2, Ype

Elgersma, PhD. Behavior and Health related Quality of Life (HRQoL) in school-aged

children with Neurofibromatosis type 1 (NF1). 12th European Neurofibromatosis Meeting,

Lisbon, Portuga~ 2001 (poster presentation).

6. L.C. Krab, F.K. Aarsen, Arja De Goede-Bolder, C. E. Catsman-Berrevoets, M. Lequin,

W.F.M. Arts, H.A. Moll, Y. Elgersma. Resultaten van medicamenteuze interventie van

leerstoornissen bij kinderen met NF1 in een dubbelblinde placebo gecontroleerde trial. In:

symposium Neurofibromatose type 1: van muis naar mens. 29e Congres Kindergeneeskunde,

Veldhoven, Nederland, 2007 (oral presentation).

7. L.C. Krab, A. De Goede-Bolder, C. Catsman-Berrevoets, F.K. Aarsen, H.A. Moll, Y.

Elgersma. Impact van NeurofibromatoseType 1 (NF1) op de schoolcarriere. 281 Congres

Kindergeneeskunde, Veldhoven, Nederland, 2006 (poster presentation).

O·L\PTF.R J(l 195

List of Publications

ARTICLES IN THIS THESIS

1. Lianne C. Krab*, Susanna M.I. Goorden*, Ype Elgersma. Oncogenes on my mind: ERI<

and MTOR signaling in cognitive diseases. Trends in Genetics, 2008;24(10): 498-510

*Authors contributed equally.

2. Lianne C. Krab, Femke K Aarsen, Arja de Goede-Bolder, Coriene E. Catsman­

Berrevoets, Willem F. Arts, Henriette A. Moll, Ype Elgersma. Impact of NFl on school

performance. Journal of Child Neurology, 2008;23 (9): In Press (doi: 10.1177/088307 3808316366)

3. Lianne C. Krab, R. Oostenbrink, Arja de Goede-Bolder, Femke K Aarsen, Ype Elgersma,

Henriette A. Moll. Health Related Quality of Life in children with Neurofibromatosis Type

1: Contribution of demographic factors, disease related factors, and behavior. The Journal of

Pediatrics, 2008; In Press (doi: 10.1016/jjpeds.2008.08.045)

4. Lianne C. Krab, Arja de Goede-Bolder, Femke K Aarsen, Henriette A. Moll, Chris I. De

Zeeuw, Ype Elgersma, Josef N. van der Geest. Motor learning in children with

Neurofibromatosis type I. Submitted (2008)

5. S.J.P.M. van Engelen, L.C. Krab, H.A. Moll, A. de Goede-Bolder, S.M.F. Pluijm, C.E.

Catsman-Berrevoets, Y. Elgersma, M.H. Lequin. Quantitative differentiation between

healthy and disordered brain matter in Neurofibromatosis type I patients using Diffusion

Tensor Imaging. AJNRAmerican Journal ofNeuroradiology 2008;29(4): 816-22

6. Lianne C. Krab, Arja De Goede-Bolder, Femke K Aarsen, Saskia M.P. Pluijm, Marlies J. Bouman, Jos N. van der Geest, Maarten Lequin, Coriene E. Catsman-Berrevoets, Willem

Frans M. Arts, Steven A. Kushner, Alcina J. Silva, Chris I. De Zeeuw, Henriette A. Moll,

Ype Elgersma. The effect of simvastatin on cognitive functioning in children with

Neurofibromatosis type 1: a randomized, double-blind, placebo-controlled trial.

JAMA, 2008;300(3): 287-294

7. Y. Elgersma, L.C. Krab, H.A.Moll. Author reply.JAMA, 2008; In Press.

194 PERSONAL NOTES

About the author

Lianne Caroline Krab was born in Velsen on July 31, 1980, and grew up in

Assendelft and Uitgeest. She obtained her VWO-degree at the Bertrand Russell

College in Krommenie in 1998, after which she attended University College

Utrecht, the international Honors College of Utrecht University, for two years,

supported by the 'Fundatie van de Vrijvrouwe van Renswoude te Utrecht'.

In 2000 Lianne started medical school at the Erasmus University Rotterdam,

and in 2001 she was admitted to the parallel Master of Science in Neuroscience trajectory. In her

traineeship in the lab of dr. Y. Elgersma she worked on the projects 'Identification of genes causing

learning disabilities in children with NF1' and 'Relationship between motor learning and cognitive

functioning in children with NF1' (in collaboration with general pediatrician drs. A. de Goede-Bolder

and neuroscientist dr. J.N. van der Geest, respectively). In 2004 she obtained her doctoral degree in

medicine, and in 2005 her Master of Science in Neuroscience degree.

In April 2005 Lianne began her PhD research on the NF1 simvastatin project, in collaboration with

the multidisciplinary CoRe (Cognitive Research) Team of the Erasmus MC - Sophia Children's

Hospital Rotterdam. During her PhD trajectory, she was co-applicant on several grants for NFl

research, and received the Jan C. Molenaar Award for the oral defense of her grant-proposal for the

scientific committee of the Sophia Kinderziekenhuis Fonds in 2005. Lianne presented her work at

national conferences and symposia, and at the international NF conferences in Portugal (with

personal funding of the Neurofibromatose Vereniging Nederland), Utah (USA), and Ireland. In

addition, she was an invited speaker at meetings of NF patient lay groups and NF1 clinical research

teams in the Netherlands, Norway, Belgium, Germany and Ireland. She wrote several contributions

for the patient magazine of the Neurofibromatose Vereniging Nederland.

In May 2008 Lianne started her medical internships at the Erasmus University Rotterdam. In 2010

she will obtain her Medical Degree, after which she plans to remain involved as a clinician and

scientist in the search for therapies for children with genetic disorders affecting cognition. She lives in

Leiden with her partner Arjen in their 1930s house which they rebuilt together.

CI-L\PTER [(I 193

Het maken van dit proefschrift was £link aanpoten, waardoor mijn vrienden en familie mij de

afgelopen jaren heel wat uurtjes hebben moeten missen. Speciale dank aan Ophirah, voor een

spoedcursus eigen prioriteiten stellen ... en nog veel meer. Pap, mam, hoewel jullie bang waren

dat ik door al dat onderzoek zou vergeten dokter te worden, zijn jullie me altijd blijven steunen.

Aan jullie heb ik mijn voelsprieten te danken, die zich heerlijk op hun plek voelen in de kliniek!

Mariska, Maurits, Fiona, we leren elkaar steeds beter kennen. Ik ben trots als zusje in onze

groeiende familie te staan. Maaike, schoonzusje, bedank voor het nalezen van mijn stukken en

de gezellige klets (vooral op zondag, heb je 'tin de gaten?).

Arjen, mijn maatje, tijdens aile (promotie)stress bleef jij als een rots in de branding, rustig,

oprecht en toegewijd. Wat bof ik dat ik jou nu al gevonden heb!

Zo, en nu eerst een ontzettend goede dokter worden!

Lianne

192 PERSONAl" NOTES

hebben we geworsteld om de l'v1RI's bij het onderzoek te krijgen. Het is gelukt! Bedankt voor

aile uitleg op mijn (zeer waarschijnlijk heel vaak dezelfde) basale radiologische vragen. Nanda,

bedankt voor al je administratieve ondersteuning.

Het Elgersma lab, jullie zijn inmiddels met zovelen dat ik het amper bij kan houden. Voor aile

leden: bedankt dat ik de pure basale wetenschap in al haar schitterende maar vaak ook harde

facetten van zo dichtbij heb mogen ervaren. Nils, ik was met veel plezier je kamergenote in ons

mini-lab met hersenplakjes achter mijn stoel. Met jou 1..-un je altijd en over alles diepgaand

discussieren. Ook manlijk advies over vrouwlijke problemen ging je niet altijd uit de weg. Wat

was het stil he, nadat ik weg ging ©. Geeske, we hebben elkaar heel wat af moeten tasten.

Enorm veel respect voor je doorzettingsvermogen en het plezier waarmee je je zo ontzettend

veellabtechnieken eigen hebt weten te maken en waarmee je je kennis aan anderen doorgeeft.

Petra, ons 3e Toppertje (in willekeurige volgorde natuurlijk), bedankt voor het klikken in de

MATLAB applicatie vanJos tot jeer scheel van werd, voor het pillen tellen tot je nagels er van

braken, en voor de lekkere babbels en cocktails. Minetta, bedankt voor het uitwerken van de

labwaardes, het meticuleus corrigeren van onze proofs, en in een ver grijs verleden je geduld met

de 'spuiten'. Thijs, met jouw enthousiasme en vlotte babbel ben je een veelbelovende opvolger

voor het NF1 simvastatine project (maar dat dacht ik al toen je voor het eerst je vinger opstak

bij de Master's, hal). Give it your best! Azar, our time together was short, but I really enjoyed

getting to know you. I wish you all the best with your fiancee!

SP 15-45, jullie zijn haast een fenomeen! Ik heb me ontzettend welkom gevoeld in jullie kleine

hokje gevuld met verhuisdozen, stingers, koekjes, pruttelende koffie en kindersurprises. Bedankt

voor de fantastische sociale en statistische input. Mirjam, je was voor mij een halve paranymf en

een hele steun. Petje af voor je heldere sociaal inzicht, en het managen van zo'n 21 keuze­

studenten tegelijk. Ik hoop dat ik mijn kinderen in de toekomst bij jou kan brengen als ze

(onverhoopt) een dokter nodig hebben. Fernke, mijn hemel wat een productie en wervelende

energie, ik twijfel er niet aan dat jij goed terecht komt. Is directrice van het SKZ niet iets voor

jou? Idse, eigenwijze nukkige lieve behulpzame (in tegenstelling tot je bewering zelden RTFM)

kerel, wanneer ga je promoveren? Ruud, zorg je als kamerjongste goed voor je roomies?

In het laatste staartje van mijn proefschrift, mede-co's: bedankt voor de gezelligheid, het delen

van aile nieuwe ervaringen, en de bekers chocomelk als ik weer eens een nacht had doorgehaald!

CHWTER 1U 191

telefoon konden kletsen over de belangrijke en totaal onbelangrijke aspecten van onderzoek en

leven. Wordt vervolgd!

Graag wil ik naast rnijn promotoren en co-promotor, prof.d:r. W.F.M. Arts, prof.d:r. B.A. Oostra,

en d:r. S.A. Kushner bedanken voor het plaatsnemen in de kleine comrnissie en het beoordelen

van rnijn manuscript, en Prof.d:r. E. Legius, R.E. Ferner, MD FRCP, en prof.d:r. F.C Verhulst

voor het plaatsnemen in de grote comrnissie. Ik ben enorm trots dat ik rnijn proefschrift mag

verdedigen voor een comrnissie die de diversiteit van mijn werk zo goed weerspiegelt. Steven,

your amazing insight and high-speed thinking generated the basis for this research. I am looking

forward to your future translational projects at our university. Prof. Arts, bedankt voor het

voorzitten van de kleine comrnissie en uw bijd:rage aan de studies in dit proefschrift. Eric, ik

hoop dat de Nederlands-Belgische samenwerking op klinisch en wetenschappelijk vlak blijft

groeien (en dat ik me bij de verdediging niet weer vergis tussen rnicro-deleties en severity?). Dr.

Ferner, thank you for coming to from London to join my promotion committee. De d:ropjes

staan klaar!

Mede NF1-auteurs, door jullie verschillende achtergronden en invalshoeken heb ik van heel veel

verschillende vakgebieden een graan~e kunnen meepikken. Jos, aan jou had ik een zeer relaxte,

positieve 4c (wat een luxe!) begeleider. Ik snap nog steeds niets van MATLAB maar gelukkig

maak je prachtige applicaties voor de blondjes. Bedankt voor al je hulp met de opstelling, de

theoretische achtergrond en de koffiepauzes op de goede momenten. Rianne, je bent een

krachtige persoon die heel goed weet wat ze wil, en we hebben menig uur* gediscussieerd over

de ideale manier van analyseren. En verhip, ondanks het feit dat je me uiteindelijk vrij liet om de

analyses naar eigen ideeen uit te voeren, zijn ze bijna zo uitgepakt als jij voorafgaand aan het

onderzoek bedacht had. Hoe doe je dat ... .? Fernke, na jaren Bourdon Vos, Rey, Beery en WISC

voel ik me haast een neuropsycholoog maar ik zie ook heel duidelijk hoeveel daar nog voor

nodig is. In zeer blijde verwachting gaat erom spannen of je er bij kunt zijn. Heel veel sterkte

met de laatste loodjes en pufjes! Coriene, jij stond altijd klaar met goede kritische feedback en

voor ruggespraak. Bedankt voor je hulp bij het interpreteren van de data (en bij de plaa*s uit de

introductie!). Marlies, bedankt voor je enorme inzet tijdens het onderzoek, onze werk- en niet­

werk gesprekken, en het actieve meedenken. J e staat er niet er voor niets (boven)op! Roell,

bedankt voor je sterke, lekker no-nonsense bijd:rage. Ik hoop dat je een super leuk onderzoek

vindt, hopelijk aan het Sophia? Saskia, dankzij jouw (zeer) geduldige uitleg werd ik een

zelfstandige SPSS prof met als specialiteit multivariate regressie analyses. Jammer dat je er niet

bij bent, maar ik hoop dat je het in de States heel erg naar je zin hebt met je gezin! Maarten, wat

190 PERSONAL NOTES

een beetje gedesensibiliseerd is (?). Met het opzetten van het nieuwe NFl simvastatine project

en onze droom van een groot CoRe centrum heeft onze samenwerking een uitdagende

toekomst!

Lieve Arja, warme, kordate, integere vrouw, alle stappen van dit onderzoek hebben we samen

genomen. AJhoewel je de credits hiervoor steeds teruggeeft, hebben we samen een brug

gebouwd van het Sophia naar de basale wetenschap, en het spoorboekje van onze trial

doorlopen. Ik kan bijna niet beschrijven hoe belangrijk jouw niet-aflatende steun de afgelopen

jaren voor mij is geweest. Hoe vaak heb jij me niet gebeld om te vragen hoe het ging? Voor mij

een eye-opener, maar voor jou overduidelijk, is dat de belangrijkste les die ik van jou kan leren

is, hoe je hart voor je werk kunt combineren met een ontspannend en vervullend persoonlijk

leven. Daamaast heb je me laten zien hoe ontzettend mooi het is om een goede, betrokken arts

te zijn, en kreeg ik de kans om te ervaren hoezeer ik mij daarin in mijn element voel. Het is voor

mij niet meet dan vanzelfsprekend dat jij op de grote dag als paranimf naast me staat. Ik ben blij

dat je, na de eerste schrik, hiertoe bereid bent.

Henriette, onze samenwerking telt een groot aantal prachtige hoogtepunten en daarmee

samengaand, zoals je ze zelf benoemde toen je mijn 'Brief 2' op de motorkap van je auto

tekende, heel wat ludieke ondernemingen. Ik ben erg blij dat ik tijdens het laatste deel van ons

project op jouw afdeling mocht werken, in een warm team onder jouw krachtige, doelmatige

(ook een toepasselijk woord) leiding. Zo heb ik een essentieel stuk basisvaardigheden voor

statistisch en klinisch verantwoord medisch onderzoek kunnen oppikken (een mens is nu

eenmaal geen muis!). Henriette, je bent een zeldzaam integer mens, doortastend en daadkrachtig,

met op alle problemen een heldere en creatieve oplossing. Geen wonder dat je met deze

kwaliteiten professor bent geworden! Ik ben er trots op jou als promotor te hebben.

Chris, bedankt voor het begeleiden van mijn promotie. J e bent een leider met visie, en je

vermogen om mensen te enthousiasmeren om het onderzoek in te duiken en keihard te werken

om er het uiterste uit te halen heb ik aan den lijve ondervonden. Ik ben benieuwd of we in het

vervolgonderzoek kinderen kunnen leren fietsen. Misschien komen we dan nog een keer op tv?

Susan, paranimf, Toppertje met een keihoog IQ en EQ, rationed met bij tijd en wijle een

lekkere portie weifelkonterigheid .... Tijdens onze labjaren volgde onze emotionele conjunctuur

gek genoeg vaak een volledig in-fase- of juist precies uit-fase patroon. Ik ben blij dat we altijd bij

elkaar terecht konden en dat we uren en uren op het lab, voor de metro-ingang, of aan de

CrL 1 J>TER W 189

Dankwoord

Bij het uitvoeren van dit promotie onderzoek ben ik rnijn sterkste, meest onvermoede en, ja,

ook rnijn minst sterke kanten tegen gekomen. Hierbij heb ik steun en inspiratie ontvangen van

heel veel verschillende mensen (bijkomend voordeel van multidisciplinair onderzoek), die

allemaal een stukje hebben bijgedragen aan rnijn groei en ontwikkeling. Hier wil ik deze voor rnij

belangrijke personen bedanken.

Allereerst gaat rnijn dank uit naar aile kinderen en ouders die mee hebben gedaan aan ons

onderzoek, voor hun hartverwarmende inzet en enthousiasme. Daarnaast wil ik de NF

Vereniging Nederland bedanken voor hun belangrijke bijdragen om het onderzoek naar de

cognitieve problemen bij NF1 te bevorderen. Zander jullie geen onderzoek!

Ype, rnijn co-promotor, zonder jou zou ik niet staan waar ik nu sta. Begin 2002 startte ik als

piepjonge Neuroscience Master student in jouw - ook nog piepjonge - onderzoeksgroep. Ik

began vol enthousiasme op het lab, maar toen ik de pipet nog steeds spuit bleef noemen, en van

hele muizenfamilies perfuseren behoorlijk ongelukkig werd, was mijn carriere bijna heel anders

gelopen. In een vlaag van de creatieve hyperactiviteit waar jouw brein om bekend staat (en die

de toon zette voor nog veel meer onderzoek) bedacht je een volledig nieuw project, waarin ik

me als een vis in het water voelde: onderzoek naar de cognitieve problemen bij kinderen met

NFL Een week later liep ik met Arja rnijn eerste NF1 spreekuur. Toen we hoorden dat jouw

vorige lab erin geslaagd was het coginitief fenotype van Nf1 muizen om te keren met behulp van

een veilige, direct toepasbare therapie, was nu of nooit. We spraken de magische woorden "dit

moeten we doen" en "dat is goed", en rnijn promotie-onderzoek was geboren (met de

bedenkers nog 'blissfully ignorant' van wat er allemaal komt kijken bij het van de grand af

opzetten van zo'n enorm project ... ).

Voor rnij was het balanceren op het grensvlak tussen basale wetenschap en kliniek een enorm

waardevolle ervaring. De contrasten tussen rationele onderzoeksminded-heid en klinisch

denken, snelle beslissingen maken en statistiek voor:if, flexibele samenwerking en comrnissies,

en 'evidence based' en 'ervaringsdeskundigheid' maakte het ons allebei niet altijd makkelijk. We

waren het lang niet altijd met elkaar eens, en onze discussies (tussen een Fries en een kwart­

Fries) zijn berucht. Ype, ik hoop dat ik een stukje van je inventiviteit en doorzettingsvermogen,

die jou zo ontzettend ver hebben gebracht, mee kan nemen in rnijn toekomstige

ondememingen. Ik hoop ook dat, in het kader van de kruisbestuiving, je witte jassen-allergie al

188 PERSONAL NOTES

vervolg trials nodig om te uit te vinden of lange-termijn behandeling met statines de problemen

op school, met gedrag en in Kwaliteit van Leven van kinderen met NF1 kan verbeteren.

186 SUMMARY

Er werd geen effect van simvastatine gevonden op de primaire uitkomstmaten (Rey Complexe

Figuur - lange termijn, de Cancellation test - snelheid, prisma adaptatie en de gemiddelde brein

ADC-waarde. In de secundaire uitkomstmaten vonden we een significant betere score in de

simvastatine groep op Figuur Leggen (~=0.54, 95% confidentie interval [CI] 0.08-1.01). Deze

verbetering was met name zichtbaar bij de kinderen met een lage uitgangs score op deze test

(~=0.80, 95% CI: 0.29-1.30). Deze verbetering in Figuur Leggen kan echter een

toevalsbevinding zijn, en de uiteindelijke conclusie van deze studie was dat een korte-termijn

behandeling met simvasatatine geen verbetering teweeg brengt in het cognitief functioneren van

NF1 patienten.

Ondanks het feit dat korte-termijn behandeling met simvastatine geen verbetering van het

cognitief functioneren van kinderen met NF1 lijkt te bewerkstelligen, vraagt het goede

veiligheids-profiel van statines om vervolg-onderzoek. In hoofdstuk 8 hebben we verscheidene

beperkingen van onze studie besproken, die mogelijk het vinden van een effect kunnen hebben

bemoeilijkt. Daarnaast hebben we aanbevelingen gegeven voor vervolg-onderzoek. De

belangrijkste aanbeveling is om kinderen voor een langere periode te behandelen. Een langere

behandelingsduur zou mogelijk het placebo-effect uitdoven, leer-effecten verminderen, geeft

ruimte voor een langere blootstelling aan de hoogste therapeutische dosis, en geeft het brein

langer de tijd om zijn functie te herstellen. Daarnaast maakt een langere behandelingsduur het

mogelijk om uitkomstmaten uit het dagelijks leven mee te nemen, zoals het functioneren op

school, gedrag en Kwaliteit van Leven. Een vervolg-trial met een behandelingsduur van 1 jaar is

op dit moment in voorbereideng aan het Erasmus MC - Sophia Kinderziekenhuis Rotterdam.

In hoofdstuk 8 concluderen we dat NF1 een grote impact heeft op het dagelijks functioneren,

omdat kinderen grote problemen hebben op school, met motoriek en, althans volgens ouders,

een lagere Kwaliteit van Leven hebben. Kennis van de problemen die kunnen v66rkomen bij

NFl kan mogelijk een tijdige herkenning en adequate interventie faciliteren. Onze onderzoeken

hebben een aantal potentiele gebieden voor structurele ondersteuning aangewezen, zoals

gedragsproblemen. Daarnaast hebben we een aantal uitkomstmaten gei'dentificeerd die gebruikt

h.'U!lnen worden op potentiele nieuwe behandelmethodes voor cognitieve problemen bij NF1

patienten te evalueren.

Alhoewel de preklinische studies hoopgevend waren, bewerkstelligt een korte-termijn

behandeling met statines geen verbetering in het cognitief functioneren. Desalniettemin zijn

CK\PTF.R9 185

adaptatie van handbewegingen. De laatste twee testen doen een beroep op de capaciteit tot

motor-leren, waarbij met name het cerebellum een rol speelt.

Hoewel we konden bevestigen dat kinderen met NFl problemen hebben met visuo-motor

integra tie en fijne motoriek, zagen we geen significante afwijkingen in de coordinatie van oog- of

handbewegingen, of in de adaptatie van saccadische oogbewegingen. Wellieten kinderen met

NFl afwijkingen zien in de prisma-gei:nduceerde adaptatie van handbewegingen. Onze

resultaten suggereren dat de problemen in de motoriek die kinderen met NFl in het dagelijks

leven ervaren deels gerelateerd kunnen zijn aan stoornissen in het motor-leren. Deze

afwijkingen lijken te worden veroorzaakt door problemen in specifieke subregio's van het

cerebrum en het cerebellum, maar niet door een volledig dysfunctioneren van deze

hersengebieden.

De meerderheid van de kinderen met NF1 laat op T2-gewogen MRl opnames van het brein

hyperintensiteiten zien, zogenaamde Unidentified Bright Objects (ongel:dentificeerde heldere

objecten; UBOs). We onderzochten de aard van deze UBOs door bij kinderen met NFl en

gezonde controles Diffusie-gewogen opnames te maken van 7 vooraf gedefinieerde

hersengebieden, waaronder de gebieden die het meest aangedaan zijn door UBOs (hoofdstuk

6). We observeerden een hogere Apparent Diffusion Content (ADC-waardes) in de

hersengebieden aangedaan door UBOs vergeleken met de hersengebieden waar geen UBO

aanwezig was, in kinderen met NF1. Daarnaast waren de ADC-waardes in kinderen met NF1

ook in de niet door een UBO aangedane gebieden hoger dan in controles. Deze verhoogde

ADC-waardes wijzen op een verhoogde totale hoeveelheid water in het hersen parenchym van

NF1 patienten. Door eigenvalues uit Diffusie-tensor opnames te onderzoeken vonden we een

indicatie dat dat deze water accumulatie zich in eerder in de myeline schede dan in axonen

bevindt.

In hoofdstuk 7 evalueerden we het effect van simvastatine op het cognitief functioneren, motor

leren en hersen-afwijkingen van kinderen met NF1 in een gerandomiseerde, placebo­

gecontroleerde, dubbel-blinde trial. 62 kinderen met NF1 van 8 tot en met 16 jaar werden

gedurende 12 weken een maal daags behandeld met simvastatine of een placebo. Het effect van

simvastatine werd onderzocht met neuropsychologische, neurofysiologische en

neuroradiologische uitkomstmaten, waarvan een deel werd gei:dentificeerd in de onderzoeken in

hoofdstuk 3, 5 en 6.

184 SUMMARY

van problemen met het functioneren op school, waarbij in ieder geval 75% van de kinderen

meer dan een standaard deviatie achterloopt bij hun klasgenoten. Daarnaast krijgt de

meerderheid van de kinderen extra ondersteuning, in de vorm van speciaal onderwijs (40%), of

remedial teaching voor (een combinatie van) problemen met leren, motoriek, spraak en gedrag

(in totaal 85%). Een belangtijke bevinding was dat de groep jonge kinderen die geen evidente

problemen in de schoolprestaties heeft, mogelijk risico loopt op het ontwikkelen van

leerproblemen, omdat zij substantiele neuropsychologische stoornissen laten zien. Tenslotte

werd er een sterke relatie gevonden tussen cognitie en de klinische ernst van NFl. Kinderen met

ernstiger klinische tekenen van NFl hadden meer problemen met het cognitief functioneren en

slechtere schoolprestaties. Deze bevindingen geven een duidelijk beeld van de grote impact van

NFl op de schoolprestaties.

In hoofdstuk 4 werd de Gezondheid Gerelateerde Kwaliteit van Leven bij kinderen met NF1

onderzocht met behulp van vragenlijsten ingevuld door ouders, en door de kinderen zelf.

Daarnaast onderzochten we de potentiele bijdrage van demografische factoren, ziekte-specifieke

factoren, en problemen met het functioneren op school en het gedrag, aan de Gezondheid

Gerelateerde Kwaliteit van Leven. Ouders rapporteerden een substantiele impact van NFl op 9

van de 13 Gezondheid Gerelateerde Kwaliteit van Leven domeinen, die problemen

weerspiegelen op het fysieke, sociale, emotionele en gedrags vlak. In tegenstelling tot hun ouders

rapporteerden kinderen met NF1 aileen problemen in het domein Lichamelijke Pijn.

Een onverwachte bevinding van ons onderzoek naar Kwaliteit van Leven was dat ondanks het

feit dat we uitgebreide schoolproblemen zagen in onze onderzoekspopulatie, deze problemen

niet lijken bij te dragen aan de Gezondheid Gerelateerde Kwaliteit van Leven scores van ouders

en kinderen. Gedragsproblemen (gescoord door leerkrachten) zijn echter duidelijk geassocieerd

met Gezondheid Gerelateerde Kwaliteit van Leven-scores van ouders. Dit wijst op een een

interessante mogelijkheid om niet aileen de gedragsproblemen zelf, maar ook de algemene

Kwaliteit van Leven van kinderen met NFl te verbeteren door gedragsproblemen aan te

pakken.

Een groot deel van de kinderen met NFl heeft problemen met de fijne of grove motoriek. In

hoofdstuk 5 onderzochten we de motorische vaardigheden van kinderen met NFl en gezonde

kinderen met een test voor fijne motoriek en visuo-motorische integratie, en paradigma's voor

de adaptatie (het leren aanpassen) van saccadische oogbewegingen en prisma-gei:nduceerde

CK\PTER 9 183

Nederlandse Samenvatting

Achtergrond (hoofdstuk 1)

Neurofibromatose type 1 (NF1) is een autosomaal dominante ziekte, veroorzaakt door een

heterozygote mutatie in het gen voor het eiwit neurofibromine. NF1 kan invloed hebben op het

lichamelijk functioneren en het uiterlijk, maar ook op het cognitief functioneren. De cognitieve

problemen bij NF1 behelzen onder andere neuropsychologische stoornissen, en problemen met

leren, gedrag, en motorische vaardigheden. Deze cognitieve problemen zijn de meest

voorkomende complicatie van NF1 op de kinderleeftijd.

Onderzoek in muizen met een heterozygote Nj1 deletie toont aan dat de het cognitief fenotype

van NF1 wordt veroozaakt door verhoogde activiteit van de RAS/ERl< signaal transductie

route. Een zeer belangrijke bevinding is dat behandeling met statines (cholesterol verlagende

middelen) de verhoogde RAS activiteit in Nf1 muizen kan terugbrengen, en hun verstoorde

synaptische plasticiteit, en problemen met leren en geheugen en aandacht kan verhelpen. Omdat

statines zo effectief zijn in muizen, en een zeer goed veiligheids-profiel hebben, zijn ze een

ideaal potentieel medicijn om de cognitieve stoornissen van NF1 patienten te behandelen.

De doelen van het onderzoek in dit proefschrift waren het verkrijgen van inzicht in de impact

van NF1 op het dagelijks leven, het identificeren van mogelijke uitkomstmaten die kunnen

worden gebruikt om het effect van potentiele therapeutische interventies te evalueren, en het

onderzoeken van het effect van simvastatine op de cognitieve problemen van kinderen met NF1

in een gerandomiseerde, dubbel-blinde, placebo-gecontroleerde trial.

In hoofdstuk 2 werd een overzicht gegeven van de huidige kennis van de etiologie van de

cognitieve problemen bij NF1 en gerelateerde aandoeningen binnen de neuro-cardio-facio­

cutane en Hamartoma syndromen, en werden potentiele behandelingsstrategieen besproken die

uit oncologisch onderzoek en uit onderzoek met diermodellen naar voren zijn gekomen.

Hoewel verscheidene studies laten zien dat NFl patienten problemen hebben met taken voor

specifieke neuropsychologische domeinen en op testen voor academische voortgang, was er niet

veel bekend over hoe deze problemen vertaald worden in het functioneren op school. Om hier

inzicht in te krijgen hebben we naast formeel neuropsychologisch onderzoek, een inventaris

gemaakt van de schoolprestaties van een grote groep kinderen met NF1, zoals beschreven in

hoofdstuk 3. We hebben aangetoond dater onder kinderen met NF1 een hoge prevalentie is

182 SUM!v1ARY

attenuate the placebo effect, wear off practice effects, enable a longer exposure to the highest

therapeutic dose, and allow for a longer time for the brain to restore function. Moreover, a

longer treatment duration would allow inclusion of real-life outcome measures such as school

performance, behavior and quality of life. A follow-up trial with a treatment period of 1 year is

currendy in preparation at the Erasmus MC- Sophia Children's Hospital Rotterdam.

In chapter 8, we conclude that NF1 has a large impact on daily life functioning, as NF1

children show substantial problems in school performance and motor functioning and, at least

according to parents, have a lower quality of life. Awareness of the problems associated with

NF1 may facilitate timely recognition and appropriate intervention. Our studies have pointed

out several potential targets for structural support, such as behavioral problems. Also, we have

identified several outcome measures that can be used to assess potential treatments for cognitive

deficits in NF1.

Although preclinical studies were promising, short-term statin treatment did not improve

cognitive functioning. Still, further clinical trials are needed to reveal whether long-term statin

treatment can improve school performance, behavioral problems, and quality of life in children

with NFL

CR\PTER 9 181

within specific regions of the cerebellum and cerebrum, but not by a ubiquitous malfunctioning

of these brain regions as a whole.

The majority of children with NF1 display hyperintensities on T2-weighed MRI of the brain, so­

called UBOs (Unidentified Bright Objects). We examined the nature of UBOs by performing

Diffusion-weighed Imaging of 7 predetermined brain regions, including those predominantly

affected by UBOs, in children with NF1 and controls (chapter 6). We observed increased

Apparent Diffusion Content (ADC values) in UBO-affected brain areas compared to UBO­

unaffected areas of NF1 children. In addition, ADC values were higher in NF1 children than in

controls, also in UBO-unaffected brain areas. These elevated ADC values indicate increased

overall water content in NF1 brain parenchyma. By examining eigenvalues obtained with

Diffusion Tensor Imaging we found evidence that this fluid accumulation is intra-myelinic

rather than axonal.

In chapter 7, we assessed the effect of simvastatin on cognitive performance, motor learning

and brain abnormalities in children with NF1 in a randomized, placebo-controlled, double blind

trial. 62 Children with NF1 aged 8 to 16 years were treated with simvastatin or placebo once a

day for 12 weeks. The effect of simvastatin was assessed using neuropsychological,

neurophysiological and neuroradiological outcome parameters, part of which were identified in

chapters 3, 5 and 6.

We did not find an effect of simvastatin on the primary outcome measures (Rey Complex

Figure test [delayed recall], Cancellation test [speed], Prism Adaptation and average brain

Apparent Diffusion Coefficient). On the secondary outcome measures, we found a significant

improvement in the simvastatin group in object assembly scores (fi=0.54, Confidence Interval

[CI]: 0.08-1.01), which was specifically observed in children with poor baseline performance

(fi=0.80, CI: 0.29-1.30). However, the results on object assembly could be a spurious finding,

and the overall conclusion of this study was that short-term simvastatin treatment does not

improve cognitive functioning in NF1 patients.

Despite the fact that short-term treatment did not reveal an effect of simvastatin on cognitive

functioning in NF1 children, the favorable safety profile of statins does call for follow-up

stu.dies. In chapter 8, we pointed out several limitations of our study that could have hampered

identification of an effect, and give recommendations for future trials. The major

recommendation for follow-up studies is to treat children for a longer period. This may

180 SUMI\iARY

found that the young children that did not have evident problems in school functioning were

potentially at risk for developing learning disabilities, as they frequently did display substantial

neuropsychological deficits. Lastly, we observed a clear relationship between cognitive

performance and clinical severity of NFl. Children with more severe clinical signs of NF1 were

more impaired in their cognitive functioning and their school functioning. In all, this study

clearly illustrates the large impact ofNFl on school performance.

In chapter 4, we assessed Health Related Quality of Life in children with NF1 using parental

reports and children's self-reports, and investigated the potential contribution of demographic

factors, disease-specific factors, and problems in school performance or behavior. Parents

report a profound impact of NF1 on 9 out of 13 Health Related Quality of Life domains,

reflecting difficulties in physical, social, behavioral and emotional aspects of quality of life. In

contrast, children themselves only reported problems on Bodily Pain.

An unexpected finding of our study on Quality of Life was that despite the fact that we found

extensive problems in school performance in our population, these problems do not seem to

contribute to Health Related Quality of Life scores of parents or children. Importantly, we

revealed that behavioral problems (rated by teachers) are a prominent predictor of parent Health

Related Quality of Life scores. This points to an exciting potential opportunity to improve not

only behavioral problems but also overall quality of life in children with NF1 by addressing

these behavioral problems.

Children with NF1 frequently display problems in fine and gross motor functioning. In chapter

5, we examined motor performance in children with NF1 and controls, using a test for fine

motor performance and visual-motor integration, and paradigms for saccadic eye movement

adaptation and prism-induced hand movement adaptation. The latter two tests assess motor

learning capacities controlled by mainly cerebellar processing.

Although we confirmed that NF1 children have problems in visual-motor integration and fine

motor coordination, we did not observe significant impairments in motor performance of either

eye or arm movements, or adaptation of saccadic eye movements. However, NF1 children did

show deficits in motor learning during prism-induced hand movement adaptation. Taken

together, our results suggest that the motor problems of children with NF1 in daily life may

partly be related to deficits in motor learning. These deficits may be caused by aberrations

CcL\PTER 9 179

Summary

Background (chapter 1)

Neurofibromatosis type 1 (NF1) is an autosomal dominant neurocutaneous disease, caused by a

heterozygous mutation in the gene encoding the neurofibromin protein. NF1 can affect physical

functioning and appearance, as well as cognitive performance. The cognitive deficits of NF1

include neuropsychological deficits, learning disabilities, behavioral problems, and motor

problems, and are considered to be the most common complication at pediatric age.

Studies in heterozygous Nf1 knockout mice have revealed that the cognitive phenotype of NF1

is caused by elevated activity of the RAS/ERK signal transduction pathway. Excitingly,

treatment with statins (cholesterol lowering drugs) can reverse the increased RAS activity in Nf1

mice, and rescue their deficits in synaptic plasticity, learning and memory, and attention. The

fact that statins are effective in Nf1 mice, combined with their very good safety profile, makes

them an ideal candidate drug to treat cognitive impairments associated with NF1 in patients.

The overall objectives of this thesis were to provide an overview of the impact of NF1 on daily

life, to identify possible outcome measures that can be used to assess potential therapeutic

interventions, and to investigate the effect of simvastatin on cognitive problems in NF1 using a

randomized, double-blind, placebo controlled trial.

In chapter 2, we reviewed the current knowledge of the etiology of cognitive deficits in NF1

and related disorders within the neuro-cardio-facial-cutaneous and Hamartoma syndromes, and

gave an overview of potential treatment options that were found using knowledge from the

oncology field and studies on animal models.

Although it was known from other studies that NF1 patients have impairments in specific

neuropsychological domains and in academic achievement tests, less was known about the

impact of these impairments on school performance. Therefore, in addition to formal

neuropsychological assessment, we inventorized school performance in a large group of NF1

patients, as described in chapter 3. We uncovered that problems in school performance are

highly prevalent among children with NF1, with at least 75% of the children lagging more than

1 standard deviation behlnd grade peers. In addition, the majority of children received additional

support, in the form of special education (40%), or remedial teaching for (a combination of)

problems in learning, motor functioning, speech and behavior (85% in total). Importantly, we

178 SUMMARY

60. Collins R, Armitage J, Parish S, Sleight P, Peto R. Effects of cholesterol-lowering with simvastatin on stroke and

other major vascular events in 20536 people with cerebrovascular disease or other high-risk conditions. Lancet

2004;363:757-67.

61. Santanello NC, Barber BL, Applegate WB, Elam J, Curtis C, Hunninghake DB, eta!. Effect of pharmacologic lipid

lowering on health-related quality of life in older persons: results from the Cholesterol Reduction in Seniors Program

(CRISP) Pilot Study. JAm Geriatt Soc 1997;45:8-14.

62. Szwast SJ, Hendrie HC, Lane KA, Gao S, Taylor SE, Unverzagt F, et a!. Association of stalin use with cognitive

decline in elderly African Americans. Neurology 2007;69:1873-80.

176 GENER.<\.L DISCUSSION AND FUTURE PROSPECTS

40. Duggan DE, Chen IW, Bayne WF, Halpin RA, Duncan CA, Schwartz MS, et a!. The physiological disposition of

lovastalin. Drug Metab Dispos 1989;17:166-73.

41. Vickers S, Duncan CA, Chen IW, Rosegay A, Duggan DE. Metabolic disposition studies on simvastalin, a

cholesterol-lowering prodrug. Drug Metab Dispos 1990;18:138-45.

42. Chen C, Mireles RJ, Campbell SD, Lin J, Mills JB, Xu JJ, eta!. Differential interaction of 3-hydroxy-3-methylglutaryl­

coa reductase inhibitors with ABCB1, ABCC2, and OATP1B1. Drug Metab Dispos 2005;33:537-46.

43. Chen C, Lin J, Smolarek T, Tremaine L. P-glycoprotein has differential effects on the disposition of stalin acid and

lactone forms in mdr1a/b knockout and wild-type mice. Drug Metab Dispos 2007;35:1725-9.

44. Thelen KM, Rentsch KM, Gutteck U, Heverin M, Olin M, Andersson U, eta!. Brain cholesterol synthesis in mice is

affected by high dose of simvastalin but not of pravastalin. J Pharmacal E>.-p Ther 2006;316:1146-52.

45. Locatelli S, Lu~ohann D, Schmidt HH, Otto C, Beisiegel U, von Bergmann K. Reduction of plasma 24S­

hydroxycholesterol (cerebrosterol) levels using high-dosage simvastalin in patients with hypercholesterolemia:

evidence that simvastalin affects cholesterol metabolism in the human brain. Arch Neurol2002;59:213-6.

46. Thelen KM, Laaksonen R, Paiva H, Lehtimaki T, Lutjohann D. High-dose stalin treatment does not alter plasma

marker for brain cholesterol metabolism in patients with moderately elevated plasma cholesterol levels. J Clin

Pharmacol 2006;46:812-6.

47. Hoglund K, Thelen KM, Syversen S, Sjogren M, von Bergmann K, Wallin A, et a!. The effect of simvastalin

treatment on the amyloid precursor protein and brain cholesterol metabolism in patients with Alzheimer's disease.

Dement Geriatr Cogn Disord 2005;19:256-65.

48. Benedetti F, Mayberg HS, Wager TD, Stohler CS, Zubieta JK Neurobiological mechanisms of the placebo effect. J

Neurosci 2005;25:10390-402.

49. McCaffrey RJ, Westervelt HJ. Issues associated with repeated neuropsychological assessments. Neuropsychol Rev

1995;5:203-21.

50. Medicines Evaluation Board. Summary of productcharacteristics ZCR-T-082004. The Hague, The Netherlands:

Medicines Evaluation Board; 2005.

51. Digilio MC, Conti E, Sarkozy A, Mingarelli R, Dottorini T, Marino B, et a!. Grouping of multiple­

lentigines/LEOPARD and Noonan syndromes on the PTPN11 gene. Am J Hum Genet 2002;71:389-94.

52. Zenker M, Buheirel G, Rauch R, Koenig R, Bosse K, Kress W, et a!. Genotype-phenotype correlations in Noonan

syndrome. J Pediatr 2004;144:368-74.

53. Kolanczyk M, Kuehnisch J, Kassler N, Osswald M, Stumpp S, Thurisch B, et a!. Modelling neurofibromatosis type 1

tibial dysplasia and its treatment with lovastalin. BMC Med 2008;6:21.

54. Wojtkowiak JW, Fouad F, LaLonde DT, Kleinman MD, Gibbs RA, Reiners JJ, Jr., eta!. Induction of apoptosis in

neurofibromatosis type 1 malignant peripheral nerve sheath tumor cell lines by a combination of novel farnesyl

transferase inhibitors and lovastatin.J Pharmacal Exp Ther 2008;326:1-11.

55. Vose LR, Arora GS, O'Brien PF, Hannan F. Pharmacologic Rescue of Behavioral Deficits in Drosophila NF1

Mutants. NF conference (abstract) 2008 p97.

56. Kotti TJ, Ramirez DM, Pfeiffer BE, Huber KM, Russell DW. Brain cholesterol turnover required for geranylgeraniol

production and learning in mice. Proc Nad Acad Sci US A 2006;103:3869-74.

57. Wagstaff LR, Mitton MW, Arvik BM, Doraiswamy PM. Stalin-associated memory loss: analysis of 60 case reports

and review of the literature. Pharmacotherapy 2003;23:871-80.

58. Muldoon MF, Barger SD, Ryan CM, Flory JD, Lehoczky JP, Matthews KA, et a!. Effects of lovastalin on cognitive

function and psychological well-being. AmJ Med 2000;108:538-46.

59. Muldoon MF, Ryan CM, Sereika SM, Flory JD, Manuck SB. Randomized trial of the effects of simvastalin on

cognitive functioning in hypercholesterolemic adults. Am J Med 2004;117:823-9.

CH.\PTER 8 175

CH.\PTER 10 I 1 9 7

198 l'ERSONAJ. NOTES